The Growth Gamble

FOR SALE & EXCHANGE
www.trading-software-collection.com
Subscribe for FREE download 5000+ trading books.
Mirrors:
www.forex-warez.com
www.traders-software.com
Contacts
andreybbrv@gmail.com
andreybbrv@hotmail.com
andreybbrv@yandex.ru
Skype: andreybbrv
ICQ: 70966433
PRAISE FOR
THE GROWTH GAMBLE
“This book is very insightful and also very practical—it provides the
practitioner with some real hands-on advice as to what to do differ-
ently. We found its contents really helped shape our thinking.”
Stephen Ford, Head of Strategy Development, Boots Group plc
"Although I do not agree with all of Campbell and Park's views,

this is an important topic addressed in a solid, fact based way,
informed by history. Few books meet these tests."
Richard Foster, author of Creative Destruction
"Campbell and Park challenge some well documented views about

growth and new business development. Their challenges are not
always successful, but they do encourage rigorous thinking about how
companies should search and select new businesses."
Robert A. Burgelman, Edmund W. Littlefield Professor of
Management at Stanford University Graduate School of Business
and author of Strategy Is Destiny
"Campbell and Park address a critical issue—new growth for the

firm. Supported by manifold examples from practice, which have
been distilled into guidelines for firms that aspire to grow new busi-
nesses, The Growth Gamble argues for a more cautious view than oth-
ers, including myself, of how much venturing a company really can
accommodate. As a result they provide a valuable antidote to exhorta-
tions for growth that drive firms to be overly ambitious."
Professor Ian MacMillan, The Wharton School and author of
Corporate Venturing and The Entrepreneurial Mindset
THE GROWTH
GAMBLE
THE GROWTH
GAMBLE
When Leaders Should Bet Big on
New Businesses and How to Avoid
Expensive Failures
Andrew Campbell & Robert Park
NIC H O L A S B R E A L E Y
I N T E R N A T I O N A L
L O N D O N B O S T O N
First published by
Nicholas Brealey International in 2005
3–5 Spafield Street 100 City Hall Plaza, Suite 501

Clerkenwell, London Boston
EC1R 4QB, UK MA 02108, USA
Tel: +44 (0)20 7239 0360 Tel: (888) BREALEY
Fax: +44 (0)20 7239 0370 Fax: (617) 523 3708
http://www.nbrealey-books.com
© Andrew Campbell & Robert Park 2005

The rights of Andrew Campbell and Robert Park to be identified as the
authors of this work have been asserted in accordance with the
Copyright, Designs and Patents Act 1988.
ISBN 1-90483-804-9
British Library Cataloguing in Publication Data

A catalogue record for this book is available from the
British Library.
All rights reserved. No part of this publication may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any means,
electronic, mechanical, photocopying, recording and/or otherwise
without the prior written permission of the publishers. This book may
not be lent, resold, hired out or otherwise disposed of by way of trade in
any form, binding or cover other than that in which it is published,
without the prior consent of the publishers.
Printed in Finland by WS Bookwell.

CONTENTS
Foreword by Professor Gary Hamel ix

Preface and Reader’s Guide 1
Chapter 1 The Challenge of New Businesses 8
Chapter 2 Beating the Odds 19
Chapter 3 The Difficulties of Building New Legs 43
Chapter 4 When Low Growth Is Better than Gambling 65
Chapter 5 The New Businesses Traffic Lights 92
Chapter 6 Diversification 150
Chapter 7 Searching for New Businesses 178
Chapter 8 Is There a Role for Corporate Venturing? 207
Chapter 9 Positioning and Supporting a New Business 231
Chapter 10 An Age of Realism 263
Appendix A Advice from Other Authors 269

Appendix B A Database of Success Stories 298
Notes 306
Index 315
Acknowledgments 322
FOREWORD
TAKING THE GAMBLE OUT

OF GROWTH
Professor Gary Hamel
S
ome wag once ventured that second marriages represent the
triumph of hope over experience. The same might be said
about the millions of dollars, euros, and pounds that compa-
nies spend every year chasing the dream of new business creation. As
any number of academic studies have demonstrated, the odds of new
business success are long—substantially longer than the odds of win-
ning big in Las Vegas.
Yet despite the long odds, companies continue to ante up. Disney
bet millions that it could build an internet portal to rival those of
Yahoo and Microsoft (MSN)—and failed. Motorola and its partners
bet millions that they could successfully scale up a satellite phone busi-
ness—and failed. Under its former CEO, Jacques Nasser, Ford badly
fumbled several attempts to acquire and scale new service businesses—
like quick-stop oil-change centers. Hoping to reduce its dependence
on the market for computer chips, Intel has, in recent years, dumped
millions of dollars into a broad array of new business ventures—and
has so far failed to produce a breakaway winner.
With new business success so improbable, one might argue that
investors would be best served if most companies returned “excess”
funds to shareholders, in the form of dividends or share buybacks. Yet
I think it is possible for just about any company to dramatically
improve its business-building odds. The fact is, there are many
TAKING THE GAMBLE OUT OF GROWTH
companies that have succeeded spectacularly in growing new busi-

nesses—despite the daunting odds. General Electric’s business-
building skills have made the company a leader in industries as diverse
as railway locomotives, engineering plastics, medical electronics, jet
engines, and financial services. Founded in 1958 to explore the com-
mercial opportunities for polytetrafluoroethylene, Delaware-based
WL Gore has created a mind-numbing array of new products and
businesses—from advanced dental floss to the world’s bestselling gui-
tar strings, from vascular grafts to membrane electrode assemblies for
fuel cells. Over the past couple of decades, Sir Richard Branson’s
Virgin Group has created dozens of new businesses and made its brash
founder a multibillionaire. More recently, Dell has extended its low-
cost, direct distribution formula into computer servers, printers, data-
storage products, and flat-screen televisions. UPS, one of the world’s
leading package-delivery companies, has recently created a thriving
business that provides logistics consulting and services. And Tesco, the
UK’s leading supermarket chain, has parlayed its distribution and
brand power into a fast-growing consumer financial services business.
IMPROVING THE ODDS
Clearly, there are some companies for whom the growth gamble has
paid off—and paid off big. More specifically, there are companies that
have learned how to consistently get more growth for less gamble. A
cursory examination of these exemplary cases reveals some simple les-
sons for improving the odds of new business creation.
LEVERAGE SOMETHING YOU KNOW HOW TO DO
First, leverage something you really know how to do. Although Disney
came a cropper with its Internet portal plans, it has succeeded in
becoming the world’s most profitable producer of Broadway and West
End shows, with award-winning productions like The Lion King and
Beauty and the Beast. Disney’s theatrical triumphs leverage the com-
x
THE GROWTH GAMBLE
pany’s deep experience in producing live entertainment for its amuse-

ment park patrons, as well as its core competence in creating memo-
rable characters and stories. Disney Cruise Lines—essentially an
amusement park at sea—is a similarly direct extension of Disney’s deep
competencies. Years ago, in a study of its new business successes and
failures, 3M discovered that new ventures based on long-standing 3M
competencies were dramatically more likely to succeed than those that
didn’t leverage core skills. Apparently, this is a lesson some companies
have yet to learn.
In Chapter 5 on the Traffic Lights, Andrew and Robert tackle this
competence issue head on under the section on “value advantage.”
They bring some much-needed rigor to the difficult issue of assess-
ing whether competencies are sufficient to give a real advantage.
DON’T BE AFRAID TO PARTNER
Second, don’t be afraid to partner. While your company may well have
some world-leading competencies, it may not possess all the assets and
skills necessary to ensure new business success. Partnerships are an
essential tool for taking the gamble out of growth. For example, in
building its financial services business, Tesco partnered with the Royal
Bank of Scotland. Virgin seldom enters a new business without a
strong partner, such as the phone giant Sprint, which is providing the
telecommunications infrastructure for Virgin’s US mobile phone ven-
ture. In bringing its trend-setting iPod to market, Apple relied heavily
on a US design partner, PortalPlayer, and a Taiwanese manufacturer,
Cheng Uei Precision. Apple is a clever company, but executives there
are humble enough to recognize the limits of its competencies. Today,
new business creation is as much about combining the competencies of
multiple companies—playing the role of orchestrator—as it is about
leveraging one’s own unique skills.
Andrew and Robert have not made partnering a central focus of this
book; my bet is that in the years ahead, alliances will become increas-
ingly critical to new business success.
xi
BE BOTH COMMITTED AND TENTATIVE
Third, be committed and tentative. It’s difficult to get a new business

design exactly right the first time. The founders of Pret A Manger, the
enormously successful UK sandwich chain, spent five years refining
their initial fast-food concept before opening their second store.
Perfecting a new business model is a process of successive approxima-
tion. Many of the early assumptions one makes around the value
proposition, ideal value chain, distribution strategy, and customer
response will, inevitably, be proven wrong in the fullness of time. The
trick is to be intellectually committed to a broad opportunity, while
being careful not to treat initial hypotheses about any particular busi-
ness model as indisputable facts.
Many managers have a hard time being both committed and tenta-
tive. The goal is to be intellectually committed to exploring a new
opportunity area, but financially prudent in doing so: to commit addi-
tional resources only as additional learning happens. Practically, this
means:
❏ being extraordinarily clear about what you do and do not know. It is

the invisible, untested assumptions that undermine success. As
Andrew and Robert so beautifully illustrate in some of their examples,
assumptions that are based on experience in existing businesses are
often the most damaging. They need special attention and testing.
❏ ranking uncertainties according to their criticality to eventual suc-
cess. Andrew and Robert propose the Confidence Check in Chapter
9 as one way of ranking uncertainties.
❏ developing creative, low-cost, low-risk ways of resolving uncertain-
ties such as early customer trials, joint ventures, prototypes, and
modeling.
❏ committing additional resources only as uncertainties are resolved.
This turns the traditional “stage gate” into a learning gate, and con-
verts an investment board into a research board, doling out addi-
tional funds only as additional learning is accumulated.
xii
THE GROWTH GAMBLE
Without an explicit learning agenda, it’s all too easy for a new business
initiative to take on a life of its own: What should have been a little
gamble soon becomes a big gamble. A surfeit of resources allows big
companies to be impatient and imprudent. “Learn more, spend less”
should be the mantra of every business builder.
EXPAND YOUR IMAGINATION AND NARROW YOUR

FOCUS
Fourth, expand your imagination and then narrow your focus. New
business creation is, at least initially, a numbers game. Take Silicon
Valley as an example. For every Cisco, Yahoo, Google, or eBay, there
are dozens of failures. That’s why the average venture capitalist invests
in a portfolio of new business opportunities rather than in a single
company. Out of a dozen start-ups, half will fail, three or four will
produce modest returns, and a couple, it is hoped, will shoot for the
moon. All too often, though, corporate business development execu-
tives try to defy these well-established odds. They surrender to the
conceit that some combination of analytical horsepower and personal
wisdom will allow them to pre-select the one or two sure winners
from among a very small number of uninspired and, all too often, rel-
atively unexamined alternatives. As is true for an ageing bachelor, or
bachelorette, a lack of attractive alternatives tends to diminish one’s
discrimination.
A far better approach is to start by throwing the net wide: make sure
you’ve developed a broad range of potential new business concepts.
Great new business ideas are few and far between. Hence, the chance
of stumbling across an outstanding business concept is directly pro-
portionate to the number of new business ideas that a company is
capable of generating. Generating a robust portfolio of truly fresh new
business concepts requires an explicit process for nurturing the mind-
sets and perceptual capabilities that have the power to reveal new
opportunities. It requires contrarians, like Michael Dell, who chal-
lenged industry orthodoxies with his original direct distribution model.
It requires seers, like the founders of Google, who saw an opportunity
xiii
in the rapidly growing profusion of uncharted websites. It requires

empathetic individuals who are able to sense the unarticulated needs of
existing and potential customers.
A company that takes the trouble to invest in building a substantial
portfolio of new business options will, inevitably, improve its odds of
coming up with a new business winner. Yet no company can success-
fully pursue dozens of new business ventures simultaneously. If a com-
pany’s new business development priorities are too numerous and too
disparate, scarce resources get fragmented and cumulative learning
becomes impossible. Luckily, a large portfolio of new business options
can almost always be categorized into a much smaller number of
opportunity domains. Having done this, a company can select one or
two growth domains on which to focus its energies in the near to mid-
term. In reconnoitering a new opportunity area, it is nevertheless
important not to overcommit to one, overly specific business model.
Indeed, it is often wise to simultaneously explore two or three business
model variants as alternate routes into an unexplored opportunity
arena.
For example Cemex, the highly innovative Mexican cement com-
pany, typically focuses its new business development efforts on a small
number of opportunity domains—such as developing low-cost housing
for the world’s poor, or developing building “solutions” for construc-
tion tractors. In this sense, Cemex is highly focused. Yet within these
domains, it is constantly experimenting with alternate business models:
developing new technologies to dramatically reduce the cost of build-
ing houses, inventing new financing options that let relatively poor
customers finance home construction, and so on. In the same way as
companies need to be concurrently committed and tentative, they must
also be simultaneously focused and experimental.
Andrew and Robert believe that a careful and meticulous evaluation
of new business opportunities will help companies lower the risk of
pouring millions down a new business rat hole. And they are right (see
Chapter 5, The New Businesses Traffic Lights). But of course, it is not
enough to merely avoid investing in a fundamentally unproductive
business concept. To create a new business winner, a company must be
xiv
THE GROWTH GAMBLE
capable of inventing novel business concepts with the power to posi-

tively redefine customer expectations, wrong-foot competitors, and
create new wealth. To take an analogy: To run a successful art compe-
tition, like the Royal Academy of Arts’ famed Summer Exhibition that
receives over 8,000 submissions each year, one needs a jury and a set of
evaluative criteria. Yet before the jury can set to work, there must first
be artists who are possessed of a creative vision. Break-the-rules inno-
vation is essential to new business success. So too is honest and
painstaking evaluation of new business options—and it is here that
Andrew and Robert make a timely and valuable contribution to our
understanding of the challenges of birthing new businesses.
In the pages that follow, you will find a wealth of deep insights,
practical advice, and meritorious admonition. I have no doubt that this
thoroughly researched and carefully argued book can help your com-
pany gamble more wisely on growth.
xv
PREFACE
& READER’S GUIDE
T
his book feels more like a destiny than a discovery. Much of
what we have been doing in the last 10–15 years has been
leading us to this.
Andrew has been researching divisionalized companies, diversifica-
tion, and corporate-level strategy for almost 20 years. During this
period there has been mounting evidence that there are limits to diver-
sification and benefits from focus. Corporate-Level Strategy, co-
authored with Michael Goold and Marcus Alexander in 1994, was
designed to help managers grapple with the issue of how much focus.
It was aimed primarily at managers in diversified companies seeking to
define a focus for their portfolios that would allow them to deliver
superior performance. This was the main challenge facing companies
in the late 1980s and early 1990s.
Corporate-Level Strategy gave guidance to companies seeking to
expand—so long as they had a “parenting formula.” The advice was to
enter or acquire other businesses that would benefit from this parent-
ing value added. Examples were companies such as 3M, Canon,
Emerson, GE, and Rio Tinto. Interestingly, 10 years later, these com-
panies are still building their portfolios based on similar sources of par-
enting value added.
However, Corporate-Level Strategy gave little explanation of why
some companies appear to be able to break the rules of Parenting
Theory. This theory, the centerpiece of the 1994 book, argued that
companies should only own businesses that the corporate parent is
good at influencing and guiding: that there should be a fit between the
skills of the parent organization and the type of businesses in the port-
folio. The problem is that this theory does not explain the many
PREFACE & READER’S GUIDE
examples of companies diversifying into very different businesses. It

did not explain how HP got from instruments to computers, how
Mannesman got from engineering to telecommunications, how GE
got from aero engines to financial services, or how Nokia got from
rubber boots to mobile telephones.
More importantly, Corporate-Level Strategy provided limited advice
to the manager whose company is maturing, who has no clear parent-
ing formula, and who is struggling to decide what to do next. As the
growth decades of the 1980s and 1990s came to an end, maturity and
the search for new businesses became more important. It was natural,
therefore, for Andrew to lead a project to examine the apparent rule
breakers and to consider what advice to give to managers in companies
like Barclays Bank, Gillette, Intel, Kelloggs, McDonald’s, Philips, and
Shell, all of whom have maturing core businesses and a desire to find
new sources of growth.
While Andrew’s research interests led to the current book, Robert’s
business activities were heading in the same direction. Through the
mid-1990s, Robert was head of strategy at National Westminster, one
of Britain’s largest banks. During these years NatWest faced some
tough strategic challenges. The core UK domestic banking business
seemed mature. Opportunities existed for expansion, both internation-
ally and into investment banking. Major corporate banking had
become a low-margin business, and investment banking looked the
way to make higher returns from NatWest’s strong corporate fran-
chise. New technologies, such as smart cards, direct delivery to con-
sumers, and the internet, offered other opportunities and threats.
Finally, managers were ambitious to become the leading UK bank by
outperforming their main rival, Barclays.
Despite many attempts (Robert counted well over a dozen in a 20-
year period), NatWest created only two new businesses of significance—
credit cards and treasury—both close to its UK banking core. In the
process the company lost hundreds of millions of pounds in failed acqui-
sitions and new ventures. So, despite the successes, the effort was not a
success overall. When NatWest announced plans to build a major life
assurance and investment business through acquiring Legal & General,
2
THE GROWTH GAMBLE
the shareholders revolted and the bank fell prey to a hostile takeover bid
from its much smaller Scottish rival, the Royal Bank of Scotland.
Robert had been uncomfortable with many of the initiatives taken
under his watch, but did not have a hard-edged way of challenging pro-
posals in the face of management enthusiasm and optimistic business
forecasts. He was therefore eager to find out if NatWest’s experience
was bad strategy, bad execution, or bad luck.
To balance out his negative experiences, he volunteered to assemble
a database of success stories. The database was biased toward improb-
able stories, like Mannesman’s move into mobile phones. It did not
contain parenting formula stories such as Procter & Gamble’s acquisi-
tion of another food business in a new geography. The database grew
to over 60 at one point, but was then culled back to around 50 as some
successes turned out to be failures and others were sold by their parent
companies.
THE RESEARCH BEGINNINGS
The project that led directly to this book started more than five years
ago as a result of a request from Hein Schreuder, vice-president cor-
porate strategy and development at DSM, a Dutch life sciences and
chemicals giant. Hein, who was a Member of Ashridge Strategic
Management Centre, pressed Ashridge to launch a project on how
companies get into new businesses. Moreover, he volunteered DSM as
a research site.
DSM has an unusual history. The company had spent most of its life
as a coal-mining business—Dutch State Mines. However, in the 1960s
the government decided to close the mines, and DSM had to find
another business to develop. In the following 10 years it successfully
transitioned to petrochemicals. This was not a leap in the dark, since
DSM had been making some petrochemicals from coal, but it was still
a remarkable transition. In the 1980s DSM developed fine chemicals
and performance chemicals businesses and was considering a second
major transition from fine chemicals into life sciences.
3
The research into DSM’s transitions was fascinating. It reinforced

many of the tenets of Parenting Theory. DSM suffered from the
difficulty of matching its parenting skills to the needs of its new busi-
nesses and hence made some costly mistakes. The research also
pointed to the importance of luck. DSM entered the petrochemical
business at the beginning of its high-growth phase. It also over-
engineered its initial chemical plants, but in a way that proved invalu-
able when managers were looking for additional volumes from
debottlenecking. The research also demonstrated the power of com-
mitment. DSM’s managers did not dillydally. Once the strategic direc-
tion was clear, managers moved forward with conviction.
The research also had its frustrations. It was not possible to assess
whether the transition had been value creating or value destroying.
DSM had entered new businesses, but had it been at the expense of or
to the benefit of shareholders? It proved impossible to tell. The com-
pany had been a state body for part of the period and had benefited
from having the rights to market gas from new fields in the North Sea
during its first transition. Nevertheless, the research sparked enthusi-
asm for more.
THE TEAM
Three other people have been particularly influential in this project.

Simon Yun-Farmbrough was head of strategy at Prudential, a UK
insurance group, at the time the project was gathering momentum. As
an ex-BCG consultant, Simon had experience outside Prudential. He
agreed to join the team meetings and share his experiences. He had
often found himself with concerns similar to Robert’s. He was skepti-
cal about some new business projects, but did not have hard-edged
tools for helping managers decide whether to take the risk or not.
Unlike National Westminster, the Prudential had, to the surprise
of the industry, launched the most successful internet bank in
Europe, Egg. Even against this success, Simon’s opinion was that
managers are frequently overenthusiastic about new businesses, take
4
THE GROWTH GAMBLE
inappropriate risks, and overlook the costs of distracting attention

from the core.
Andy Morrison had been closely involved with new ventures at BG,
Britain’s largest independent gas company. He had been part of the
team that identified and selected new-venture projects. He had helped
some projects through their development stages. He was much more
positive about the value of new businesses.
In the research team, Andy took on the challenge of identifying cor-
porate venturing units that had successfully developed significant new
businesses for their parent companies. This involved interviews with
seemingly successful corporate venturing units. It also involved linking
up with Professor Julian Birkinshaw at London Business School, who
was running a parallel project on corporate venturing.
To Andy’s surprise, although not to Simon’s, he failed to find any
convincing examples, even when the research was expanded to include
corporate venturing units set up in the 1980s. Corporate venturing
units have their uses, he and Julian concluded, but they do not gener-
ally help companies to develop significant new businesses.
Klaus Volkholz had been head of strategy at Philips’s Dutch head-
quarters from the mid-1980s to the mid-1990s and was now based in
California. He was also an enthusiast for new businesses and corporate
ventures. He had had considerable experience in both throughout his
career at Philips. He had been part of successful and also failed efforts
to create new businesses and managed a corporate venturing effort from
start to closedown. He agreed to join the team to help us expand our
US-based research, to provide some balance to a growing skepticism
about the value of investments in new businesses, and to share with us
some of the lessons from Philips’s experience over the previous 30 years.
With regard to Philips, Klaus summarized 42 significant invest-
ments to get into new businesses over three decades, which are part of
the public record. Five made major contributions to shareholder
value—Integrated Circuits, ASM Lithography, Polygram, Navteq, and
BSO/Origin—providing us with some challenging examples to test our
theories against; 15 other initiatives, mostly part of divisional new busi-
ness efforts, were clearly successful additions to the Philips portfolio.
5
Of all new business initiatives, 18 are still part of Philips, while 24 have
been discontinued or sold, including some of the successful ones.
Klaus judged that the total expenditure on new businesses probably
outweighed the value of the successes. This was no great surprise, since
the period from 1970–1990 was not the most glorious in Philips’s his-
tory. Many of the new businesses, particularly in computers and com-
munications, suffered from being confined to European national
organizations because of the group’s organizational structure. While
Philips’s overall record was extra food for the skeptics in the team,
Klaus pointed out that in high-tech industries there is an imperative to
develop new businesses to exploit emerging technologies. His under-
standing of this imperative and his personal belief in the value of the
opportunities in new technologies contributed enormously to balanc-
ing the team.
The material that follows is, therefore, not just the work of Andrew
and Robert. We take the blame for any failings and omissions, but need
to give credit to those who have helped our journey.
HOW TO READ THE BOOK
This book is for a mix of audiences: senior managers, business develop-

ment executives, consultants, academics, and students. Each group may
want to take a different path through the book.
We suggest senior managers read Chapters 1 and 2, which describe
the growth challenge, using Intel and McDonald’s as examples, and sum-
marize the main messages. Based on their interest, they can then read
any or all of Chapter 5 (tool for screening new business proposals),
Chapter 7 (advice on what processes to set up to increase the flow of
good projects), Chapter 8 (the value of corporate venturing units), and
Chapter 9 (how to look after and control a new business after its early
beginnings).
Chapter 4 (When Low Growth Is Better than Gambling) is a chap-
ter we would like all managers to read, because it helps them under-
stand that growth is not an absolute requirement. It is a matter of
6
THE GROWTH GAMBLE
choice. This understanding will help managers be wiser when they

start looking at new businesses.
Business development executives will probably want to start with
Chapter 5 (the screening tool), Chapter 8 (advice on corporate ventur-
ing units), or Chapter 9 (advice on how to look after a new business as
it develops). They will then enjoy reading Chapter 3, which explains
why so many mistakes are made, and Chapter 4, which shows them
why their job is not indispensable.
Academics will want to take a different path through the book.
Many will want to dive straight in at Chapter 5, to understand the
screening tool we have developed, or Chapter 6, which places this tool
in its academic context. The latter chapter discusses, in particular, the
advances we think we have made over previous theory. Academics will
also want to browse through Appendix A, in which we compare our
views to that of a number of other prominent thinkers in the field.
Consultants will want to spend a good deal of time on Chapter 5, to
understand the screening tool, and Chapter 9, to understand how to
support a business as it starts out. For those with a little more time, a
gradual browse through the book from the beginning will pay huge
dividends in helping consultants position themselves with the client. In
the area of helping companies with growth and new businesses, many
consultants do their clients more harm than good. This book has many
subtle differences from current received wisdom that can only be fully
absorbed by following through all the chapters.
Students will want to get their bearings with Chapters 1 and 2 and
then focus on Chapters 5 and 6, which contain the main intellectual
work. They will then want to read Appendix A to find out more about
what other writers have said. If they have time, they will be well
rewarded by reading Chapter 3, which explains why failure is so com-
mon, and Chapter 4, which explains why growth is optional.
We hope you enjoy your read.
7
CHAPTER 1
THE CHALLENGE OF
NEW BUSINESSES
I
magine you are a senior manager at McDonald’s or Intel. Both
companies have had very successful histories. Intel was started in
the 1970s as a producer of integrated circuits for memory prod-
ucts. In the 1980s it focused on microprocessors and by 2000 the com-
pany was worth nearly $400 billion. McDonald’s was founded in the
1950s. Since then it has become the world’s best-known fast-food
chain, and in 2000 had grown to a market capitalization of over $60
billion.
Both companies face a tough challenge: the future does not look as
good as the past unless you, as one of the senior management team,
can help find some new avenues for growth. Fortunately, both com-
panies are positioned in industries that are growing. The demand for
both semiconductors and good-value restaurants is expected to con-
tinue to expand. But the segments that Intel and McDonald’s are
strongest in—microprocessors and hamburgers—are weak. In fact,
recent performance has been poor. In 2003 McDonald’s published its
first quarterly loss for 40 years and Intel’s net income for 2002 was a
third of its peak. Furthermore, both companies have devoted signifi-
cant resources to the quest for new growth businesses, with little suc-
cess to date. So as a senior manager at either company, what should
you do? Should you launch a number of small initiatives? Should you
take a big gamble on some major acquisition or greenfield invest-
ment? Should you be more cautious?
THE GROWTH GAMBLE
THE INTEL STORY
Intel has made many attempts in the last 10 years to enter new busi-
nesses. Some have been successful, but most have not. Fortunately,
Intel’s efforts have been carefully studied by one of the most well-
regarded academics on the subject of new business development—
Professor Robert Burgelman—and summarized in his book Strategy Is
Destiny.1
Intel began life in the 1970s using integrated circuits to produce
electronic memory products. It was the early leader in this market as a
result of superior technology and design skills. As the memory market
exploded, Intel began to lose share to more aggressive competitors
focused on volume rather than superior technology. However, as a
result of a chance request from a Japanese customer, Intel began
designing microprocessors. As margins in the memory market began to
fall, more attention was given to higher-margin and higher-technology
product areas like microprocessors. By the early 1980s, Intel had
become a microprocessor company in terms of profits and technical
commitments, even though the chief executive, Gordon Moore, still
described it as a memory company. In fact, it took a change in leader-
ship from Moore to Andy Grove to complete the transformation and
to exit the memory business.
Intel then entered a golden period of around 15 years during which
its dominance of the microprocessor market for PCs and servers pow-
ered the company’s growth and extraordinary profitability. During this
period, there were many attempts to enter new businesses. Some were
intended to help extend the core microprocessor business into new
areas, while others concerned growth opportunities outside the core
business strategy. Intel managers even labeled these two types of ini-
tiatives Job 1 (extending the core) and Job 2 (beyond the core). In prac-
tice, it was not easy to define objectively whether a new initiative was
Job 1 or 2. But the way the effort was treated inside Intel depended
strongly on whether top managers viewed it as Job 1 or 2.
In 1993 Frank Gill took charge of Intel Products Group (IPG), the
organization unit with responsibility for most of the Job 2 initiatives. It
9
THE CHALLENGE OF NEW BUSINESSES
included new businesses such as motherboards for PCs, LAN adapter

cards, fax modems, PCMCIA cards, video conferencing, and massively
parallel supercomputers. Gill, as recorded by Burgelman, was trying to
create a significant new business for Intel. “My main challenge in com-
ing to IPG was to find opportunities for Intel to develop a significant
business without competing with our customers.” One of Gill’s subor-
dinates, Jim Johnson, explained: “Andy [Grove] wants a second prod-
uct line to balance microprocessors.”
By 1999, however, little progress had been made. Burgelman
concludes:
Despite the emergence of many new opportunities across the comput-

ing industry, and the emergence of many new ideas within the com-
pany, Intel had had difficulty turning these opportunities into
successful stand-alone businesses. … Grove enumerated some fifteen
non-microprocessor initiatives and observed that only chipsets, moth-
erboards, networking and Intel Capital had been successful.
In fact, his criteria for success were limited. Only Intel Capital had
earned a decent return on invested money. Intel’s lack of success was in
part because the core microprocessor business was growing so fast and
was demanding so much support that initiatives without a justification in
terms of Job 1 were viewed with suspicion. Despite the stated intention
of growing a significant new business, the follow-through was weak.
This caused Craig Barrett, Grove’s successor, to refer to the micro-
processor business as a “creosote bush,” comparing it to the desert plant
that weeps creosote onto the ground, killing anything that grows near it.
Another reason for the run of failures was the difficulty Intel had in
managing businesses with a different business model. The successes
tended to be close to microprocessors and “predictable.” Intel’s cul-
ture, functional structure, and managerial rules of thumb were driven
by the needs of the microprocessor business, with its largely pre-
dictable technological progress. In businesses with different rules and
technical judgments that depended on insights about the future, Intel
struggled. Frank Gill commented:
10
THE GROWTH GAMBLE
While the functional organization worked great for our core micro-
processor business, it was very difficult for my group who depended on
the functional groups for support.
Overall, Andy Grove concluded that he had failed in his ambition to

develop a significant new business. After he had passed the CEO’s
mantle to Craig Barrett, he admitted to an MBA class at Stanford:
The old CEO knew that this [slowdown in the microprocessor business]
was coming. He tried like hell to develop new business opportunities, but
they almost all turned into shit.
With Craig Barrett came a new energy to push Intel into new busi-
nesses. His first action was to define a direction of development. Intel’s
mission, he decided, was to change from being a microprocessor com-
pany to being an internet building block company. The focus was
switched from computers to the internet. The urgency was reinforced
by a clear slowdown in sales growth. Revenues in 1998, the year Barrett
took over, grew by only 5%.
THE MCDONALD’S STORY
McDonald’s has had around 40 years of exceptional performance.

Founded in 1955, the company was driven forward by Ray Kroc,
whose passion was value for money and consistency. McDonald’s went
public in 1965 and powered ahead of its rivals by opening restaurants
fast. The company then drove growth by going international. By 1995,
nearly 50% of its operating income came from outside of North
America, and in 1997 70% of new restaurant openings were outside
North America.
By 1998, McDonald’s operated 25,000 restaurants in 115 countries.
In that year Jack Greenberg became only the fourth CEO in
McDonald’s history. In his first annual report he was able to point to a
10-year record of 21% compound annual total return to shareholders.
11
However, he was worried about the company’s ability to continue its

long performance record. Growth potential in the core hamburger
business was becoming limited, its strong position in the US market was
being threatened, growth opportunities in international markets were
reducing, and customers were raising concerns about beef and/or fatty
foods. In fact, financial results in 1996 and 1997 had not been satisfac-
tory. Quinlan had opened his 1997 shareholders’ letter with an apology:
“1997 was a disappointing year.”
Greenberg’s first initiative was to focus on reviving performance in the
US with the launch of Made For You. This required some investment in
every US restaurant, a special cost that hit 1998 figures. He also reduced
the number of new restaurant openings from around 2,500 per annum to
fewer than 2,000. At the same time, he began to look for opportunities to
enter new businesses. In 1998 he acquired a minority stake in Chipotle
Mexican Grill, a 14-restaurant chain in Denver, as part of a strategy to
“create additional franchising opportunities for our McDonald’s fran-
chisees.” In 1999 he acquired Donatos Pizza, a 143-restaurant chain in
the US, as an opportunity to “learn about how McDonald’s competencies
in real estate, restaurant operations, marketing, … [and] unique global
supply infrastructure can be leveraged.” He also acquired a small coffee
chain with some 30 sites in the UK, Aroma Café.
Jim Cantalupo, CEO of McDonald’s International, explained the
strategy in terms of McDonald’s share of the quick service restaurant
(QSR) industry. These new brands enhance “our ability to serve more
customers across the QSR industry.”
We owe it to our shareholders, owner/operators, employees and sup-

pliers to carefully explore selected, complementary opportunities to
capture more of the total meals-away-from-home market.
In 2000 and 2001 a number of other initiatives followed:
❏ The acquisition, from Chapter 11 bankruptcy, of Boston Market, a

chain of nearly 900 restaurants, some with potential for converting
to McDonald’s or to its Donatos or Chipotle Partner Brands.
12
THE GROWTH GAMBLE
❏ An investment in Food.com, a company known for its internet food

takeout and delivery service.
❏ A joint venture with Accel-KKR to launch eMac Digital, a vehicle
for creating trading exchanges and business service platforms that
leverage McDonald’s strengths.
❏ A minority stake in Pret A Manger, a UK sandwich chain.
❏ The testing of two new McDonald’s concepts: McDonald’s with a
Diner Inside, providing table service and a 122-item menu, and
McCafe, offering gourmet coffee and related products.
❏ Finally, in 2002, McDonald’s signed a letter of intent with Fazoli’s, a
400-site fast-casual Italian restaurant, to jointly develop the concept
in three US markets. The letter of intent also gave McDonald’s an
option to purchase the company.
This rash of initiatives was partly boosted by a decision in 1999 to give

a remit for new growth initiatives to the then strategy director, Mats
Lederhausen. He now became responsible for new businesses and for
stimulating innovation (excluding Partner Brands). His thoughts
included the following:
We are very humble about our growth challenge. We realize how hard
it is to identify new paths of growth. I wish we had more answers to the
question you are asking but I don’t. All we know is that we have to try
to keep growing. We know that the misery of uncertainty is far better
than the certainty of misery. If we don’t try to grow we will fail, and at
least if we try we have a chance of sustained growth.
The pipeline of new businesses must be dynamic and vibrant. There

must be initiatives at different stages of development. When a big busi-
ness hits a stall point it takes time to develop new initiatives. So these
projects must already be in the pipeline. This does not mean that you
should let 1000 flowers bloom. I hate this. But it does mean that you
should not let the pipeline dry up.
13
RECENT HISTORY AT INTEL AND MCDONALD’S
In early 1999, Craig Barrett decided to focus Intel’s commitment to

new businesses. This was in response to a groundswell from executives
who had been on an internal management program titled “Growing
the Business.” This program had labeled Intel’s microprocessor busi-
nesses as “blue” businesses and all new businesses as “green”
businesses.
Barrett put Gerry Parker, one of two executive vice-presidents, in
charge of the green businesses under the heading New Businesses
Group. The New Businesses Group contained one big initiative, Intel
Data Services, Intel’s entry into the web-hosting business, and a num-
ber of smaller ones. “We’re focusing on two areas, the internet and
appliances,” explained Parker.
However, despite setting up the New Businesses Group and Craig
Barrett’s decision to free up his own time by appointing Paul Otellini
“in charge to a large extent of the core business,” Intel’s efforts were
still not solving the problem. Burgelman comments that, in early
2001, when John Miner was taking over the responsibility for new
businesses from the retiring Gerry Parker, “top management seemed
to have concluded that it was difficult to develop businesses that were
not related to the core business.” Moreover, the “internet building
block” vision was now being toned down to “PC Plus,” implying that
Intel would in the future focus on businesses that were closer to its
PC/microprocessor heartland.
In 2003, Intel announced that it was withdrawing from web hosting
and wrote off around $1 billion. However, undaunted by the continu-
ing failures, in 2004 it committed to another bold set of diversification
moves. The focus was chips, but the ambition was to dominate addi-
tional segments of the chip industry. Intel had over 80% of the market
in PCs and servers. Its ambition was to enter and dominate the mar-
kets for chips in flat-panel displays, handhelds, personal media players,
Wi-Max, a new wireless technology, and cellular phones. At the time
of writing, progress on these ambitions is mixed. In August 2004, Intel
announced that it was holding back its plans to enter the projection
14
THE GROWTH GAMBLE
television business because of delays in its next-generation computer

chip. Working models of the Pentium 4, the next generation, created
too much heat. Moreover, Intel was losing market share and technical
leadership to its smaller rival, Advanced Micro Devices.
In early 2005 Paul Otellini, due to replace Craig Barrett in May,
announced a reorganization around user platforms. “The new organi-
zation will help address growth,” he explained. At the same time,
Advanced Micro Devices issued a profits warning due to aggressive
pricing by Intel in flash memory products.
McDonald’s most recent experience has been somewhat similar.
Between 1999 and 2002 the core hamburger business performed even
less well than expected, and in 2002 Jack Greenberg was replaced by Jim
Cantalupo. In the reshuffle, Mats Lederhausen, already in charge of
strategy and innovation, also became responsible for Partner Brands.
The message from McDonald’s was that greater management atten-
tion would be focused on improving the core business and that contin-
uation of some of the Partner Brands and other new initiatives would
be reviewed. Dow Jones reported that McDonald’s had contacted sev-
eral leveraged-buyout firms and financial sponsors to gauge their inter-
est in buying a 51% stake in the Partner Brands portfolio (Boston
Market, Chipotle, Donatos, Pret A Manger, and Fazoli’s).2
In January 2003, McDonald’s announced its first ever quarterly loss
as a result of plans to close 700 underperforming restaurants and cut
costs. Cantalupo also declared that the historic gospel of 10–15%
growth was no longer sustainable. Instead of opening new stores, he
would focus on “getting more customers in our existing stores.” The
back-to-basics medicine worked. By 2004 McDonald’s was announcing
double-digit growth in its US business and planning to revitalize its
European business.
At the time of writing, McDonald’s still retains some of its new busi-
nesses, but its ambitions have been reduced. Pret A Manger, for exam-
ple, abandoned most of its international expansion plans, changed
some of its top managers, and announced plans to refocus its energies
on the UK market. Along with the four other remaining new busi-
nesses, Pret has been placed in McDonald’s Ventures. Mats
15
Lederhausen, managing director of McDonald’s Ventures, has been

charged with deciding whether any of these businesses can become
significant for McDonald’s without distracting from the core. For
those that cannot achieve this difficult objective, Lederhausen will
develop an exit strategy.
If McDonald’s exits all of the businesses in McDonald’s Ventures, it
will have failed to develop any significant new businesses despite a 10-
year effort. However, its search for growth has not been a failure over-
all. The core business has revived. The new focus on healthy menus
was in part stimulated by consumer insight work carried out in the
search for new growth. The retailing of DVDs at McDonald’s stores is
an idea that came from the portfolio of new initiatives, as is the Wi-Fi
access that is now available at many sites. According to Lederhausen,
the difficulty McDonald’s had with its new businesses made managers
realize that there was unlikely to be any alternative to reviving the core.
This helped in the revival. Moreover, Lederhausen is still hopeful that
some of the Partner Brands will become significant new businesses.
McDonald’s has also had to cope with unexpected deaths. Jim
Cantalupo died of a heart attack in April 2004. His successor, Charlie
Bell, was diagnosed with cancer weeks after taking over and died in
January 2005. No organization can face two such public tragedies
without being deeply affected.
WHY DID THEY FAIL?
To many, the failures at Intel and McDonald’s seem most peculiar.

Surely McDonald’s, a powerhouse of fast-food retailing, must have
many opportunities. Its skills in supply chain management, franchising,
branding, and international expansion must be of enormous value to
any new restaurant concept looking to expand. Yet it has consistently
found it difficult to make a growth success out of its portfolio of
Partner Brands; and Pret A Manger, one of its most successful Partner
Brands, appears to be rejecting McDonald’s help and refocusing on its
core.
16
THE GROWTH GAMBLE
Moreover, Intel and McDonald’s are not alone. According to Business

Week, Microsoft has invested billions in its new businesses—games,
MSN, business software, and software for mobiles and handhelds—with
little success in terms of growth or profit.5 Or consider Kodak, a classic
case of a 20-year effort to get into digital photography that has until
recently been plagued by failure. The oil companies have also tried many
other businesses with little success. In fact, the experiences of McDonald’s
and Intel are the norm. Despite huge resources, despite management
commitment, despite taking big risks, large companies with maturing
core businesses find it very hard to develop new growth businesses.
Klaus Volkholz, a member of our research team, long-time watcher of
the semiconductor industry, and an ex-head of strategy at Philips, found
Intel’s failures particularly hard to understand. He pointed out that Intel
has strengths in manufacturing and design, which should be a huge
advantage in many of the adjacent semiconductor businesses. He specu-
lated that the problem must be Intel looking for businesses that earn
returns as good as microprocessors—a hurdle that is impossible to leap.
We were so surprised by Intel’s failures that we located a manager who
had first-hand knowledge of the Intel experience. He explained that the
failures, as seen from the inside, were even more significant. He pointed
out that it was not from lack of trying. In his view, the problem was the
management mindset baked into every Intel manager. This mindset is
the microprocessor mindset: a virtual monopoly business, where para-
noia about competitors, refusal to double source, technical arrogance, a
lack of customer sensitivity, and many other mental habits dominate.
These habits work in the peculiar environment that is the microproces-
sor business. They are just plain wrong for most other businesses.
We even observed the ramp-up at one of Intel’s new Fabs (fabrica-
tion plants). Everything about Intel’s unique way of doing things was
confirmed. Managers were working battlefield hours with a total ded-
ication to driving yield and making the Fab the best in the Intel fam-
ily. But, we could see that these habits—the work hours, commitment
to yield regardless of cost, fanaticism about safety, and so on—might
well be a disadvantage in other businesses where Intel has less control
of the market.
17
It is this observation, in Intel and in many other companies, that

gives us our cautious position about new businesses: the strengths of
the core business are often weaknesses when applied to other, even
adjacent businesses.3 Moreover, these strengths and weaknesses are so
ingrained in the minds of managers that they are not easily changed,
even when you take the managers out of the company.
Take Paul Allen, co-founder of Microsoft. He has had the same dif-
ficulty developing new businesses outside Microsoft as Bill Gates has
had inside. Business Week reported that he has “burnt through” $12 bil-
lion following a vision of a “wired world.” His idea was that the tech-
nologies lying behind the convergence of entertainment and
communications would be gold mines.
But the strategy crashed along with the tech boom, its demise com-
pounded by Allen’s missteps. Often he paid wild prices for harebrained
schemes. Other times he bought too soon.4
We believe that companies like Intel and McDonald’s can only find sig-
nificant new businesses under two circumstances. Either they discover a
business that “fits with” their existing businesses and responds well to the
habits and rules of thumb that apply to the core. Or they go through a cri-
sis that breaks their commitment to the old habits and rules of thumb, and
brings in new leadership and new ideas at the top and in the middle.6
IBM is an example of a company that has transitioned from its core
twice. Both times required wrenching changes. The first occasion, the
move to computers, was driven by a unique manager, Thomas Watson
Jr, at war with his father. The second time, the move to services, IBM
had to go through one of the biggest crises in US industrial history. We
rarely see companies solving the problem with the middle-ground
solution that many authors propose: more innovation, more support
for new businesses, and more separation of new ventures from the
core. Hence, since this book is not primarily about how to engineer a
crisis, it is aimed mainly at helping managers at companies like Intel
and McDonald’s find those few businesses that will fit with their exist-
ing core.
18
CHAPTER 2
BEATING THE ODDS
I
ntel and McDonald’s illustrate an important reality: most compa-
nies fail to find new growth businesses when the core businesses
mature. Over 15 years for Intel and nearly 10 for McDonald’s,
these companies have not found a significant new business that will
enhance their future growth prospects. And they are not alone.
If we take the most pessimistic statistics, as many as 99% of compa-
nies fail to create successful new growth platforms. The Corporate
Strategy Board’s study of growth restarts concludes that less than 10%
of companies manage to restart growth once it has slowed. Only 3%
sustain a restart for more than three years. Less than 1% do so by cre-
ating new growth platforms.1
Yet nearly every company tries. There are only a tiny percentage of
management teams who settle for sticking to their core businesses and
declining gracefully as the business matures. Chris Zook, a consultant
from Bain & Co., points out in Profit from the Core that 90% of com-
panies aim for growth rates more than twice that of the economy, while
less than 10% succeed for more than a few years.2
This is why we have labeled the challenge of finding new businesses
a growth gamble. To even take on the challenge is to bet against the
odds. This is also why you will find a rather sober approach to the chal-
lenge in the following pages. We will not be suggesting that there are
any easy answers. You are not going to be advised to set up a corporate
venturing unit or to let 1,000 flowers bloom. You are not going to be
told that the solution lies in creating a more innovative culture. In fact,
you are going to learn that much of the received wisdom is dangerous.
You will discover that entering new businesses is highly risky and
should be avoided unless the circumstances are clearly favorable. You
BEATING THE ODDS
are going to be encouraged to focus on existing businesses first. You

may end up less enthusiastic and more cautious about entering new
businesses. You will understand why the odds are against you and learn
when they are likely to be in your favor. You may even decide that
McDonald’s or Intel or your company would be better advised to stick
to its existing businesses for the next few years rather than risk major
investments in new areas. Importantly, you will be armed with some
tools for helping your company make the right decisions. You will be
able to help it enter new businesses when the circumstances are right
and avoid them when the circumstances are wrong.
As you read on, you should remember that we are talking about new
businesses, not new products. This book is about the challenge of cre-
ating new legs for a portfolio that is beginning to mature, not new
products or new markets that extend the franchise for an existing busi-
ness. Sometimes this distinction, between new businesses and exten-
sions to existing businesses, is blurred. McCafe and McDonald’s with a
Diner Inside are examples on the borderline. Our definition of a new
business is “a separate business unit that has a business model different
from that of the existing businesses.” Go, the low-cost airline, was a
new business for British Airways. Keebler, the cookie company, was a
new business for Kelloggs. While this book may help managers
wrestling with extensions to existing businesses, it is primarily aimed at
those looking for new businesses that can be grown into whole new
legs to the company’s portfolio.
A SHORTAGE OF OPPORTUNITIES THAT FIT
Much of the current literature argues that the high failure rate of new
businesses is due to poor processes and skills.3 Companies are advised
to become more entrepreneurial, to copy some of the approaches used
in the venture capital industry, and to build a process for developing
new businesses.
Our view is different. We have observed that established companies
have established managerial habits, rules of thumb, and mindsets.
20
THE GROWTH GAMBLE
These mindsets are normally well tuned to the needs of the existing
businesses, but get in the way when the company tries to enter new
businesses. Success comes when companies select new businesses that
respond well to these mindsets. Failure is often the result of trying to
do things that do not fit. Moreover, for many companies there are very
few opportunities that fit. In other words, the main difficulty that man-
agers face is the fact that there are likely to be very few good opportu-
nities for their company to enter new businesses.
This shortage of opportunities explains the low success rate and
implies a different way of approaching the problem. It suggests that
efforts to generate additional ideas and experiment with a portfolio of
new ventures are likely to be fruitless. It suggests also that a screening
tool that helps managers identify opportunities with a reasonable fit is
likely to be more useful than a series of process steps for developing
new businesses. In fact, trying to make dramatic changes to managerial
mindsets, as suggested by those authors who encourage companies to
set up a separate new businesses process, is also unlikely to succeed.
Short of changing many senior managers and challenging the com-
pany’s ways of working—the crisis solution—the inherited mindsets
and rules of thumb will continue to influence success.
We recognize that our view can be criticized for being defeatist.
However, we will show that it is not. Managers who face up to the real-
ity of “few significant opportunities that fit” will find they are better
armed to help their companies succeed than those who believe that
more experimentation and more new ventures are the answer. But first
we need to share our research data.
OUR RESEARCH
We shadowed for a number of years managers in companies like BG,

the major British gas company, McDonald’s, and Shell. In every case,
we concluded that these managers were investing in too many projects,
most of which had little chance of success. In one case 24 significant
ideas were identified and examined and 11 were launched as new
21
BEATING THE ODDS
ventures. However, when put through a screening process developed

by the research team only one small venture passed the screen; two
others were marginal. In the three years since, eight of the projects
have been closed down, some with hundreds of millions in losses. The
remaining three are either small or still high risk and, in aggregate, do
not make a profit. In other cases there were many more new initiatives
(one manager referred to 361 proposals his department had processed),
but only a few passed our screen.
Our second research avenue involved assembling a database of suc-
cesses. It is evident from this database that most successes are a result
of careful strategic choice rather than multiple experiments or port-
folios of initiatives. Of the sample of more than 50 successes, which was
biased toward examples of diversification rather than extension, more
than 70% originated from a process of strategic planning: thinking
through the options and choosing a few new businesses to invest in.
Centrica, a British retailer of gas that had previously been part of a
nationalized industry, is a good example. Faced with new competition
in a previously protected market, Centrica forecast tighter margins and
reduced volumes. This spurred managers into a major review of what
the company should do. The review was divided into an upstream proj-
ect focused on gas supply and storage and a downstream project
focused on retailing and customer relationships. Many new business
opportunities were considered, such as call center outsourcing building
on the company’s experience servicing its 17 million customers.
The review concluded that the company should expand its gas asset
interests, enter the business of retailing electricity, explore retailing
telecoms, and look for acquisitions of companies with business models
that involved a mobile workforce and large call centers. The latter
thought led the company to acquire AA, the UK motor rescue and
financial services business. While some of these strategies have not
progressed as well as hoped, the overall result turned a company with
declining prospects into a top-quartile performer.
Hewlett-Packard’s move into computers, IBM’s success in business
consulting, and GE’s development of financial services are also exam-
ples of decisions that were driven by careful strategic choice. In all
22
THE GROWTH GAMBLE
three cases, the company already had some activity in these new areas
that could be built on: in-house sourcing of CPUs at HP, an unloved
consulting and services activity at IBM, and customer financing at GE.
However, these seeds were not sown deliberately. They were resources
these companies could build on that, at a certain point in time, pre-
sented a good opportunity and fit with the company’s managerial
approach.
None of the examples in the database started out as part of a cor-
porate incubator, as a result of an internal program to create a “third
horizon,” as a cultural change effort aimed at generating new busi-
nesses, or as a deliberate program of investing in a portfolio of prom-
ising new ideas. About 20% of the examples were more opportunistic
than planned. For example, Royal Bank of Scotland’s decision to invest
in Direct Line, described in more detail in the next chapter, was clearly
opportunistic. Peter Wood was looking for a company to back his idea.
He happened to know the manager who had become finance director
of Royal Bank of Scotland. One day, when he was “playing golf” with
this manager, he suggested that Royal Bank support the project.
Twenty years later Direct Line is one of Royal Bank’s most successful
businesses.
Our third research avenue involved surveying, together with
Professor Julian Birkinshaw of the London Business School, the suc-
cesses and failures of corporate venturing units. These units operated
under different rules from their parent companies. In fact, they were
often set up to avoid the mindset constraint of the existing businesses.
During the second half of the 1990s, most companies launched one or
more corporate venturing units or corporate incubators. Normal rules
of corporate risk aversion were suspended. Corporate funds and third-
party venture funds were available for all promising projects. Many of
these units mimicked the processes and methods of the venture capital
industry. In other words, they were following current advice. They
were investing in more projects, taking more risks, and trying to create
the seeds that would lead to significant new businesses.
Yet out of a sample of more than 100 units, the number of signifi-
cant new businesses that was created for the parent company was fewer
23
BEATING THE ODDS
than five. Moreover, it was apparent that the total costs far exceeded
even optimistic estimates of the value created. Our conclusion was that
these units were not an appropriate way of trying to tackle the growth
challenge. It turns out that corporate venturing does have a place, but
not as a process for developing new growth businesses.4
SIX RULES FOR NEW BUSINESSES
In contrast to the advice that fills most of the pages of most books and
articles, we have concluded that it is the shortage of significant oppor-
tunities that fit that is the biggest challenge for managers seeking new
growth. This has led us to some different recommendations. To guide
managers in companies like Intel and McDonald’s, we have developed
six rules about new businesses.
CONTINUE TO INVEST IN THE CORE
The first rule is to continue to invest in the core businesses. Writing in

a business column for the Independent on Sunday titled seductively “My
biggest mistake,” Sir Peter Walters, a former CEO of British
Petroleum, described attempts by his company to enter new business
areas. Driven by concerns about the future of the oil industry follow-
ing the 1970s oil shock, BP entered nutrition, IT, minerals, and other
businesses either because of the similarity of the skills required (min-
erals) or because of their growth potential (IT software). After describ-
ing the failures, Sir Peter concluded:
If we had put even half the effort into our core businesses that we put
into new businesses, we would have come out ahead.
We have heard the same story again and again from companies that
have lost faith in their core businesses. John Steen, a manager from
British fibers company Courtaulds, described it as a “strategic round
trip.” He explained:
24
THE GROWTH GAMBLE
You decide that the core is not growing fast enough. You invest in new
businesses. They underperform; and seven years later you are worse off
than when you started having closed or sold the last of the new initia-
tives. I have seen my company do this three times.
In Profit from the Core, Chris Zook has made much of this point.5 Yet
most companies have still not got the message. Oil companies, tobacco
companies, bulk chemical companies have all tried to find new busi-
nesses to invest in. For most the costs have been high and the benefits
hard to find. Moreover, during their campaigns to diversify, they have
given too little attention to their core businesses. As Sir Peter pointed
out, the real cost of investing in new businesses may be a distraction
from the core, rather than money lost from a failed venture. Moreover,
if there are no new business opportunities that fit, the company’s future
will depend on the performance of the existing businesses. Unless
management’s first priority is to maximize the potential from these
businesses, the future challenge may be survival rather than growth.
What we are suggesting is that senior managers need to have their
feet firmly on the ground when they decide in which new businesses to
invest. One sure way of shortening the life of their company is to invest
unwisely in new businesses.
For many companies patience is a virtue. For a period of years, it
may be best to focus on their core businesses, even if they are not
growing fast. At intervals during this period the company will want to
review its opportunities for new businesses and it will want to keep an
opportunistic watch at all times. However, it may take 5, 10, or 15 years
for the right opportunities to come along. Trying to force the pace can
speed rather than stem decline.
Kelloggs spent 20 years on many fruitless initiatives before it found
a way, with the acquisition of Keebler, of moving successfully beyond
branded cereals into convenience foods. Crown Cork and Seal, which
we will describe in more detail in Chapter 4, is an example of a com-
pany that kept focus on its core for 40 years while its competitors
chased new businesses in the evolving packaging industry. Their new
businesses created poor performance and led them first into the hands
25
BEATING THE ODDS
of break-up specialists and then into the arms of Crown. Often a com-
pany can survive longest by focusing on its existing businesses and
exploiting the occasional opportunity to step into new businesses.
We say more about focusing on the core in Chapter 4.
DON’T BE SEDUCED BY SEXY MARKETS, BUT RECOGNIZE

RARE GAMES
The second rule is to beware sexy markets. In one company, the head
of the corporate development division was focusing on telecommuni-
cations and internet opportunities because these were the only indus-
tries that were growing fast enough to allow for the creation of large
amounts of value in a short period. The result was more than $600 mil-
lion in write-offs.
In another company, the focus was on diversification into
biodegradable products. Growth in this sector was expected to accel-
erate as legislation tightened. A suitable acquisition was identified by
the business development team, but rejected by a cautious board.
Subsequently, the target company was forced to close when too many
competitors entered the market.
The attention given to growth markets is understandable. Managers
are looking for growth. But so are all their competitors. To create
value, managers should be focusing on markets where they have an
advantage, rather than markets that are growing. In fact, it may be eas-
ier to create value in less popular areas than in “businesses of the
future” that attract every growth-hungry manager. What managers
should look for are opportunities where their company can bring some
special resource or competence to the game.
There are only two exceptions to this advice: “dog markets” and
“rare games.” Dog markets are ones where most competitors currently
earn less than their cost of capital or are likely to do so in the future.
Airlines and steel are good examples, as were semiconductors in the
1980s. They normally occur because the competition has become too
intense. Dog markets should be avoided whether the company has
advantages or not.
26
THE GROWTH GAMBLE
Rare games are markets where even average competitors are likely
to do well. They normally occur because a new market suddenly opens
up, creating more demand than supply. Those companies that get in
early have an advantage in terms of written-down assets or stronger
brands. Internet service provision and fiberoptic networks were exam-
ples in the early 1990s.
Another reason for rare games is the behavior of competitors.
Sometimes they price high, making room for new entrants. Sometimes
they are so hampered by legacy processes and habits that unencum-
bered competitors have an advantage. These conditions have existed in
financial services in many countries, allowing new entrants such as GE
Capital and even supermarkets to enter the business and do well.
A company can enter a rare game without an advantage because it
is an easy environment. To be sustainable, however, the company needs
to create an advantage during the first few years. This normally comes
from early-mover benefits.
We say more about how to identify dogs and rare games in Chapter
5.
LOOK FOR ADVANTAGE, DON’T PLAY THE NUMBERS GAME
Many authors argue in favor of creating a portfolio of new business ini-

tiatives. The logic is as follows. Investing in new businesses is risky.
Failure rates are high, with estimates of 80%, 90%, and even higher
numbers being quoted. Statistics from the venture capital industry are
often used. To get one significant success, companies need to screen
1,000 business plans, investigate 100, and invest in 10. Gorilla Park, a
new business incubator, is an example. In the two years up to the end
of 2001, the incubator had looked at 2,000 proposals and invested in
16. As Kirby Dyess, a former senior Intel executive, put it: “In order to
get a home run, we need to go to bat multiple times.”
It is argued that companies need dedicated venturing units in order
to process this volume of work and manage the large number of invest-
ments that need to be made. Moreover, since the one or two successes
will take a number of years to emerge, companies need a pipeline of
27
BEATING THE ODDS
new initiatives. In the same way that companies have a new product
development process and dedicated staff, they should also have a ded-
icated new businesses development process. In fact, we started our
research with this hypothesis.
We now believe that the numbers game is a losing game. Our third
rule, therefore, discourages managers from having multiple tries and
from even thinking with this mindset. If the presumption is of many
failures and few successes, each new initiative will get back-of-the-
hand support from the organization because it will be expected to fail.
A numbers game mentality is not only a bad gamble, it contributes to
failure rather than success.
The alternative is to be selective: to invest in opportunities only
when the company has a significant advantage. Sometimes this may
mean no new initiatives. At other times it may mean two or three. But
for a particular level in the organization, it should never be 10 or 12 or,
as in one company, 44.
Take the analogy of new product launches. Every marketing man-
ager knows that the failure rate is high. Yet each new product launch
has the full support of the organization. As a consequence, businesses
rarely launch more than two or three new products a year. If you are
going to give something your full support, you cannot do it to more
than two or three at a time. This forces managers to weed through the
ideas for the very best, and then to try to make them work.
In opposition to this selection thesis are authors who point out the
difficulty of picking winners. If the venture capital companies cannot
pick winners, why will Intel or McDonald’s be able to? Moreover, there
are many stories of success by accident and of successes with counter-
intuitive business models that would be unlikely to pass any screen.
Having studied many new business projects and shadowed many
managers in the process of choosing which projects to support and
which to reject, we believe that wise selection is possible.
So, when does a company have a significant advantage? The answer
is when it believes it can serve the market and earn 30% better margins
than competitors. When Rentokil, a company in pest control, consid-
ered entering the office plants business, its advantage was its business
28
THE GROWTH GAMBLE
model. Managers believed they could charge a 30% premium by offer-

ing a higher-quality service. All other things being equal, this would
translate to a 200% increase in profits.
When Kelloggs considered buying cookie company Keebler, man-
agers knew that they could drive significant additional volume through
Keebler’s distribution system: they could add Nutrigrain and other
products to Keebler’s portfolio. In addition, they knew that Keebler’s
strong direct sales and merchandising force would increase the sales of
these Kelloggs products. All other things being equal, these changes
could add 50–100% to Keebler’s profits.
However, for most new business projects, the 30% challenge is too
great. Either an existing competitor has many of the same advantages
or it is possible to imagine other likely competitors with similar advan-
tages. When British Airways launched its low-cost airline Go, it had
some advantages. The parent company could give the new business
some landing slots and help with operational support. But when com-
pared to established competitors such as Ryanair and easyJet, these
advantages would not have met the 30% hurdle. Go struggled as a new
business under BA’s wing, did better as a management buyout, and
ended up being acquired by easyJet. Optimists can persuade them-
selves that they have a 10 to 20% advantage, but a 30% hurdle is high
enough to require more than optimism.
We say more about the numbers game in Chapter 3 and about
advantage in Chapter 5.
BE HUMBLE ABOUT YOUR SKILLS
The fourth rule is to be humble about the current skill set, at least with
respect to the new business. One of the reasons the advantage hurdle
needs to be set at 30% or greater is because managers are frequently
overoptimistic about their skills. The Rentokil or Keebler advantages
only turn into 100% increases in profits if all other things are equal.
Normally, however, all other things are not equal. Competitors
have advantages that are difficult to observe. Moreover, there are
learning costs in a new market. In unfamiliar situations, managers
29
BEATING THE ODDS
make mistakes: they invest too much in marketing, overdesign prod-

ucts, mismanage suppliers, and more. Acquisitions are often used to
overcome learning costs. But acquisitions involve premiums. This is
why the advantage for the new company needs to be at least 30%. If
the learning costs or the premiums are likely to be high, the expected
advantage needs to be even higher, 50% or 100%, to make sure that
there is a significant residual advantage.
It is the learning costs that explain why, for any specific company,
there are few opportunities. This is obvious in sport. We would expect
a team skilled at soccer to perform pretty poorly playing American
football, and we would expect the Olympic table tennis champions to
fail to qualify for the tennis event. Yet we often expect similar leaps of
skill in business.
Learning costs are an unknown that upset many plans, but they can
also suggest one type of new business idea to look out for—saplings.
Saplings are operating units that already exist within the company. They
are often unloved or ignored, but exist for historical reasons or reasons
connected with the core businesses, which have the potential to be
grown into new legs for the company. Their special feature is that they
have, for whatever peculiar reason, especially strong managers with
insights about how to grow the activity. Their other big selling point is
that most of the learning has already been done.
Hewlett-Packard first started manufacturing processors for its
instrument business in the 1960s because it did not want to rely on
third-party suppliers. In the 1970s the unit started selling mini-
computers for technical applications. Around 1980, after many
requests, the unit’s strong management team was allowed to enter the
market for commercial applications. Now, after the addition of servers,
software, and personal computers, HP is one of the world’s largest
computer manufacturers.
Hotel company GrandMet acquired the IDV wines and spirits busi-
ness in 1974, along with its bid for a brewer called Trumans.
Management tried to sell IDV to pay down debt, but failed. Twenty
years later, under the guidance of a uniquely talented manager,
Anthony Tennant, IDV had become the largest business in GrandMet.
30
THE GROWTH GAMBLE
The most dramatic example of a sapling in our database is IBM

Global Services. Now accounting for nearly 50% of IBM’s revenues, and
in itself the world’s largest computer services company, it began life in
the late 1980s as a sapling. Some maverick vice-presidents persuaded a
reluctant IBM to respond to requests by some major customers such as
Eastman Kodak to bid for their IT operations on an outsourced basis. A
small arm’s-length operation was set up, which developed its own life
well away from the core business, and was largely ignored until Louis
Gerstner became IBM CEO at a time of crisis in 1994. Even then he left
IBM Global Services well alone for two years, waiting until he had
started to turn around the core hardware and software businesses before
bringing the growing services business center stage.
By understanding the impact of learning costs, managers can discover
good prospects for new businesses where much of the learning has
already been done, as well as avoid potential disasters where the learning
cost is likely to be larger than any advantage the company possesses.
We say more about learning costs in Chapter 5 and about saplings
in Chapter 6.
SEARCH FOR PEOPLE AS MUCH AS POTENTIAL
The venture capital industry has a saying that there are only three
things to think about when selecting projects to support: management,
management, and management. The same applies to new businesses
inside larger companies.
Large companies are lulled into presuming that somewhere, within
the huge pool of talent, managers can be found to lead projects. The
challenge, they believe, is to find good projects for them to lead. While
this approach works within an existing business, it does not work when
the challenge is new businesses. Within an existing business there are
many managers who understand the products, the markets, and the
essence of the business model needed to make a profit. In new busi-
nesses all three of these areas of knowledge may be absent.
Exercises that focus on identifying opportunities can be dangerous.
They build enthusiasm for some new opportunity before the
31
BEATING THE ODDS
management that can exploit the opportunity has been identified. Often
this leads to projects starting out with the wrong leadership. A company
may have a huge advantage in technology or brands or customer rela-
tionships, but if it does not have the managers with the talent, experi-
ence, and will to exploit these advantages, no value will be created.
In all of our research sites management was an issue. In some, like
Egg, the internet bank created by Prudential, management was the key
to success. Mike Harris, hired in to run the bank, was about the only
manager in Britain who had had the experiences needed to make a suc-
cess of the opportunity. He had previously run First Direct, Britain’s
first major direct-to-consumer bank. He had also been involved in
other technology-driven start-ups.
In other research sites, management was the scarce resource. “We
had great ideas, but we did not have managers with the right entre-
preneurial skills to exploit them” was a common theme. In one case,
the company launched two new businesses and the head of business
development became chief executive of both. A year later this manager
was moved to run a larger division. In another company, the chief exec-
utive confided that his failure to succeed in mobile telephones was
greatly affected by not having a “strong enough leader” for the busi-
ness. He explained:
It is not just a matter of technical talent. You need someone with the
will to win, someone who can get the rest of the organization to sup-
port the project. These people are rare.
It is the rarity of these people that drives the need to focus on people
as much as potential. Any process to search for new business ideas
should be complemented by a search for managers with the experience,
talent, and passion to lead. In fact, since these managers are likely to
have their own ideas about what sorts of new businesses they want to
create, it is often better to start with a search for talent rather than a
search for opportunities. As Larry Bossidy, president and CEO of
AlliedSignal, puts it: “At the end of the day, you do not bet on strategy,
you bet on people.”
32
THE GROWTH GAMBLE
We say more about identifying opportunities and the importance of

people in Chapters 5 and 7.
BE REALISTIC ABOUT AMBITIONS
Our sixth rule is to set ambitions in the light of opportunities. There is

a macho attitude in management thinking that argues for setting high
targets. In Built to Last, Jim Collins refers to BHAGs (Big Hairy
Audacious Goals).6 Managers, he argues, achieve more dramatic
advances when they set audacious goals. Leaders of companies, like Intel
and McDonald’s, who are facing a big growth challenge are advised to
set a big goal: “Create a billion-dollar business within five years” or
“Raise our growth rate from 10% to 15% in the next three years.”
Our view is that the successes managers have had using stretch goals
to power performance in their existing businesses cloud them from
realizing that the same techniques cannot be used to drive the creation
of new businesses. Stretch goals work in existing businesses because
managers get stuck in ruts and stretch goals can unlock their thinking.
Moreover, their knowledge of the existing businesses enables them to
be wise about which breakthrough ideas are foolish and which are
gems.
In the search for new businesses, managers have no ruts to break out
of. Moreover, they have much less instinct about which new ideas are
acorns and which are rabbit droppings. Stretch goals just distract them
from thinking rationally. Hence, in our view, goals should be set only
after the opportunities for new businesses have been screened, not
before.
One director of corporate development was asked to generate $500
million of operating profits from new businesses in five years. He made
the target more manageable by setting himself the objective of creat-
ing 10 new businesses, each with $50 million of profit. But the target
pushed him into risky investments that resulted in $500 million of
write-offs rather than $500 million of new profit. While this was a par-
ticularly bad outcome, similar stories emerged from most of the other
companies.
33
BEATING THE ODDS
The need to avoid setting ambitious goals has major implications

for the way that efforts to get into new businesses should be organized.
If managers create a new businesses division or even a development
unit, pressure will mount from bosses and subordinates to set targets
and define objectives. This was an issue in most of the companies shad-
owed in the research. Where targets were set, they were more of a hin-
drance than a help, interfering with objective assessment of the
opportunities rather than stimulating additional ideas. The task should
be to identify opportunities and see if any of them are worth support-
ing—to explore the potential rather than fill a growth gap. The task of
exploring new businesses is, therefore, likely to be better done as a
project team with a finite time frame rather than by setting up an orga-
nizational unit.
In addition, managers need to be prepared to discover that there are
no new businesses that the company should invest in at the moment:
that the project team may come up with a blank. Fear of this outcome
is one of the reasons managers set up new businesses divisions. The
expectation is that there will be a stream of new projects or acquisitions
to invest in. Our research suggests that this is irrational. Most compa-
nies only have a few new opportunities that fit their capabilities. Hence
there is no need for an incubator or new businesses division.
The alternative—giving money back to shareholders—may feel
defeatist. But it can still be the right thing to do, at least until an oppor-
tunity that is likely to succeed emerges. Moreover, it is possible to pro-
duce acceptable performance for shareholders so long as growth
performance is not significantly lower than expectations. Over the last
10 years IBM has grown in line with the economy, but, by buying back
more than 25% of the company’s stock, it has produced a total return
to shareholders in the top quartile.
We say more about setting targets and organizing the new busi-
nesses effort in Chapter 7.
34
THE GROWTH GAMBLE
Figure 2.1 The New Businesses Traffic Lights
THE STRATEGIC BUSINESS CASE
We believe that managers can use the power of basic strategic analysis
to help them find those opportunities that fit. At Ashridge Strategic
Management Centre we have developed a screening tool: the New
Businesses Traffic Lights (see Chapter 5). The tool can be applied to
an idea before a business plan has been developed. It can also be
applied alongside a business plan to assess the strategic logic for the
proposal. It can even be applied to an existing investment that is fail-
ing to meet its short-term targets to see if the strategic logic is still
sound.
The Traffic Lights are a distillation of good strategic thinking.
There are four lights: the size of the value advantage, the attractiveness
of the profit pool, the quality of the managers running the new ven-
tures and their corporate sponsors, and the likely impact on existing
businesses (see Figure 2.1). Each element is scored red, yellow, or
green. Red implies that we have a significant disadvantage, or that the
market segment has little available profit, or that our managers and
sponsors are inferior to those of the competitors, or that there is a big
negative for existing businesses. Any one red light signals that the
35
BEATING THE ODDS
probability of success is too low, even if some of the other lights are
green.
Green implies that we have a big advantage, or that the market seg-
ment is an easy one to make money in, or that our managers and spon-
sors are clearly superior to competitors in this market, or that there are
big benefits for existing businesses. Any one green light (without any
red lights) suggests that the probability of success is favorable: the
company should invest so long as a viable business plan can be
developed.
Situations that have all yellow lights are marginal. Ideally, managers
should be asked to reformulate their proposal so that at least one light
can be green. Alternatively, an experimental investment can be made in
the hope that experience will show that one of the yellow judgments
should be green.
Applying this screen to a typical portfolio of new business invest-
ments results in red lights for many projects. Our conclusion is that
most companies are taking too many risks in search of new growth
businesses. Either they are driven by a commercially irrational concern
about survival, or they have been encouraged to make too many invest-
ments by the current literature. The potential for improvement is
huge. Not only can significant money be saved from the “new busi-
nesses” budget, but extra resource can be focused on improving the
core businesses, an activity that is likely to give good rewards. By
reducing the amount they gamble, managers can improve their exist-
ing businesses and keep some powder dry until better projects emerge.
HOW OUR ADVICE DIFFERS FROM OTHER

AUTHORS
There are several excellent books on the subject of growth into new
businesses (see Appendix A for our analysis of the advice these books
give).7 None takes exactly the same focus as this book—achieving
growth by developing significant new businesses, especially in compa-
nies whose core is slowing down. However, they all throw some light
36
THE GROWTH GAMBLE
and some confusion on our topic. The difficulty is that many of the
authors are addressing innovation and growth in general, including
innovation to improve the core, innovation to extend the core, innova-
tion to create significant new businesses, and innovation to create
smaller new businesses. This confusion between innovation in general
and the creation of significant new businesses is, we believe, the cause
of some misunderstandings about our topic that need to be corrected.
We have already commented on two points of disagreement with
the current literature: the advice that managers should play the num-
bers game, and the suggestion that companies need to set bold targets
for their new businesses effort. In addition, we want to alert managers
to two other messages in the literature that can easily mislead the
unwary.
1 Success is about creating a supportive environment for

entrepreneurs, for “purposeful accidents,” and for exploring the
periphery
A common theme in the literature is about releasing potential.

Corporate cultures are viewed as risk averse and blinkered; good ideas
are killed off before they get a chance to develop. If companies could
only create the sort of supportive environment that occurs in Silicon
Valley, growth problems would surely be solved.
This view of the world presumes that there are many good oppor-
tunities and many managers with latent entrepreneurial skills. The
solution is to run “ideation workshops” to tease out the natural cre-
ativity and knowledge in managers. An internal market needs to be
created between entrepreneurs and providers of funding that mirrors
the external marketplace. Systems need to be changed to promote and
reward risk taking. Cultural change through training and selection is
required.
In our view, there are very few situations where good ideas and good
entrepreneurs combine with good operators and good luck. This is
because there are few good ideas, fewer good entrepreneurs, and a lim-
ited amount of luck. Hence the probability of it all coming together is
37
BEATING THE ODDS
low. To some degree this is why so many successes seem more acci-
dental than planned. The late 1990s, when conditions were most
favorable to entrepreneurs with ideas, spawned many thousands of
would-be Bill Gates, but did not greatly increase the number of suc-
cesses. In the same way, the suggestion schemes that many companies
have tried can generate a large number of ideas without creating much
additional value.
We therefore believe that most good projects (the few “accidents”)
will shine brightly enough to be identified by managers using appro-
priate screening criteria. Thus companies need to put more effort into
the process of top-down screening: eliminating all but the most excep-
tional ideas, entrepreneurs, and “accidents.” In our experience, at least
half of the new business projects that companies launch could be
screened out before any significant investment is made.
This rational view contrasts somewhat with the more emergent and
even mystical view taken by some. Mats Lederhausen of McDonald’s,
for example, stated:
I believe the mysticism surrounding growth is why there are so many

books written about it and why so few seem to get it right. Growth is
situational, it is contextual, it is philosophical and yes it is even spiri-
tual. The quest for growth is a marathon not a sprint. It is a journey not
a destination. It is culture not strategy. Ultimately it is less about where
you want to go and more about who you are.
Mats was referring to growth in general rather than new businesses in

particular, but also argues that the distinction is less than useful. He
points out that McDonald’s new business activities have succeeded.
They did not create a significant new business, but they contributed to
a rebirth of the core business and McDonald’s is now growing again—
the main objective in the first place.
We acknowledge McDonald’s turnaround of the core business, but
are committed rationalists. More rationality and less attention to
preparing the conditions for “purposeful accidents” will, we believe,
bring better results.
38
THE GROWTH GAMBLE
In addition, human beings are biased in favor of positive thinking

and optimism. This is especially true of managers. A philosophy
toward new businesses that is excessively negative is no more use than
one that is foolishly hopeful. Managers need to believe that they will
find new businesses that they can exploit. However, they need to search
for these opportunities recognizing that the task is not like seeking a
needle in a haystack, implying that you will find it if you look hard
enough, but is more like looking for the lights of a rescue ship when at
sea. Sometimes when you look there are definitely no lights, and some-
times you think you see a light in the distance but are not sure. When
finally a ship does appear, the lights are unmistakable. If you wait for
this moment, you will not have used up your flares on mirages, and you
will have plenty of energy left to attract the ship by ringing bells,
shouting, and waving.
For many, the analogy of looking for the lights of the rescue ship is
much too passive. We agree. However, it is a useful contrast to the des-
perate search we see many companies involved in. Using the needle in
the haystack mindset, they examine every piece of straw, even taking
some of the stronger pieces to see if they will act as needles. Another
benefit of the sinking ship analogy is that this is often how it feels for
managers, who become increasingly worried that the core business is
sinking under them. Like ships, existing businesses normally sink much
more slowly than managers predict. Moreover, holes below the water-
line can often be repaired. The rule, therefore—with sinking ships and
with core businesses—is to focus first on making them as seaworthy as
possible and to abandon them only when you are confident you have
something safer to board.
2 Success is about learning the disciplines of new business

creation and integrating them with the disciplines of running
the core businesses
Some companies are continual creators of new businesses—new boat

builders! Famous examples are 3M, Corning, Canon, and Virgin.
These companies have a track record of spawning new businesses again
39
BEATING THE ODDS
and again, and they have management processes, competencies, and a

culture that support this. In addition, in Silicon Valley and elsewhere,
there are venture capital companies and technology incubators whose
whole reason for existence is to nurture and develop new businesses.
Many authors have therefore concluded that the answer to the new
businesses challenge is for companies to learn the skills possessed by
3M or the venture capitalists. They observe that these skills are rather
specialist and are often very different from the skills of running
McDonald’s or Intel. Hence they argue that companies need to
develop these skills in parallel with those required to run the core busi-
nesses. Over time, they should meld the two sets of skills together into
an integrated management approach that will allow the company to
drive hard in its existing businesses and, at the same time, innovate and
nurture new businesses.
In our view this is unrealistic. There are so few examples of compa-
nies that succeed in achieving the combination of continuous new busi-
ness creation and dedication to the core that this is, for most, an
unrealistic objective. Some companies devote all their energies to new
business creation and, as a result, often exit businesses that were previ-
ously core—tapes for 3M, music for Virgin, consumer products for
Corning. Most, however, focus on driving their core businesses and
successfully adding new businesses only when the needs of the new
business fit with the management approach that is required to run the
existing businesses.
Moreover, our database of successes demonstrates that companies
can successfully add significant new businesses to an existing portfolio
without developing and integrating the skills of continuous new busi-
ness development. In fact, many of our successes would not be consid-
ered to possess the skills of new business creation that are so frequently
described in the literature. Admittedly, some have been lucky. But
most—like GE in financial services, Mannesman in mobile telephones,
or Prudential in internet banking—made a strategic decision to expand
in a certain way and were successful.
Rather than advise companies to develop the skills for continuous
new business development, we believe it is more realistic to encourage
40
THE GROWTH GAMBLE
them to focus on a less ambitious objective. They should maintain a

watchful eye for promising new businesses through careful selection
processes, and then focus their efforts on making a success out of the
few promising opportunities that can gain the commitment of the
organization.
GAMBLING ON NEW BUSINESSES
All companies need a strategy for new businesses. In this book we are
focusing particularly on a subset of companies—those, like
McDonald’s and Intel, that have a growth challenge. These two com-
panies are concerned that their core businesses are running out of
steam, and they feel the need to find major new revenue streams. We
feel confident that the advice we have for companies like McDonald’s
and Intel is also relevant for others. But we are cautious. Our research
has focused on companies that have successfully entered significant
new businesses or are trying to do so.
Many authors present the world as one of plentiful opportunity.
They see a changing world with opportunities in many directions: new
types of demand, new technologies, and new business models.
Companies, they argue, need to create a supportive environment for
entrepreneurial activity, learn new skills, stretch their imagination, and
try a number of new initiatives.
Our research leads to a different conclusion. We believe there are
few significant opportunities: it is a world of scarcity rather than plenty.
Significant opportunities are rare because so many factors need to fit
for the probability of success to be high: a large market, advantages
over competitors, managers capable of exploiting the market and the
advantages, and minimum disruption to the existing businesses.
Therefore our advice is more about caution than encouragement, more
focused on selection than experimentation, and more linked to
patience than activity.
The lack of opportunities and the inherent risks make the growth
gamble difficult. However, it is made more difficult than it need be
41
BEATING THE ODDS
because of the way many approach the challenge. Some of this is the
fault of managers, who tend to overcommit to unrealistic objectives
and overestimate their capabilities. But some is the fault of consultants
and academics, who have given managers well-intentioned but mis-
leading advice and encouragement. We say more about this in the next
chapter.
We hope this book will help managers develop more realism in their
strategies for entering new businesses. We hope it helps them avoid
taking risks when the odds are stacked against them and encourages
them to invest with confidence when the odds are in their favor.
Overall, managers can gamble and beat the odds, but only once they
fully understand the game they are playing.
42
CHAPTER 3
THE DIFFICULTIES OF
BUILDING NEW LEGS
W
e have established that most companies fail when they try
to enter new businesses. Moreover, when companies
gamble and fail, it is often at huge cost both to share-
holders and to management.
Clayton Christensen, author of The Innovator’s Dilemma and The
Innovator’s Solution and one of the world’s top strategy gurus, uses
AT&T to illustrate the problem. During the 1990s AT&T, America’s
leading telephone company, lost more than $50 billion trying to get
into new businesses. It made three big attempts: computers with NCR,
mobile telephony with McCaw Cellular, and cable broadband with
TCI and MediaOne. They were all disasters.
Christensen concludes:
We could cite many cases of companies’ similar attempts to create new-

growth platforms after the core business had matured. They follow an
all-too-familiar pattern. When the core business approaches maturity
and investors demand new growth, executives develop seemingly sensi-
ble strategies to generate it. Although they invest aggressively, their
plans fail to create the needed growth fast enough; investors hammer
the stock; management is sacked; and Wall Street rewards the new
executive team for simply restoring the status quo ante: a profitable but
low growth core business.”1
The same statistics are recognized by every author on the topic. Intel
reputedly lost $1 billion on its web-hosting business. Even companies
THE DIFFICULTIES OF BUILDING NEW LEGS
that succeed often do so at the expense of long-term shareholder value.

Richard Foster and Sarah Kaplan, authors of Creative Destruction and
deep thinkers about long-term company survival, provide chilling
data.2 They identified 18 companies that remained in the Forbes top
100 between 1917 (the first year the data was compiled) and 1987.
These remarkable survivors, companies like General Electric, DuPont,
and Procter & Gamble, might be expected to be the exemplars other
companies should copy. However, Foster and Kaplan conclude the
opposite. As a group, these survivors underperformed the market by
20% over the 70 years. They may have survived but, as a group, they
destroyed value.3
In order to understand why managers often destroy value when they
try to build new legs, we shadowed executives responsible for helping
their companies find new businesses. Over a period of four years we
shadowed 10 managers. Shadowing involved speaking to the manager
every three months or so to find out what issues he or she was working
on and what successes and failures had occurred in the previous period.
Most of these efforts did not survive the time of our research and only
two of the efforts survived the full period of our shadowing. Three of
the managers were encouraged to leave their companies as a result of
failure. The research, however, gave us invaluable insights into the
problems managers face and the mistakes they make.
In addition to the shortage of opportunities, described in the previ-
ous chapter, there are two reasons investments in new businesses often
destroy more value than they create. First, managers are trying to do
something that is inherently risky. Like prospecting for oil, a high fail-
ure rate is to be expected. This occurs because each project contains
many risks that cannot be managed away. This source of failure makes
any decision to enter a new business a gamble, at least in part.
The second reason for failure is more avoidable. It stems from the
way that managers approach the task. Without a firm theoretical
framework for thinking about new businesses, managers find them-
selves following guidance and thought patterns that cause them to
make mistakes. The fault lies primarily with the managers responsible
for the new businesses effort, but it is also partly shared with their
44
THE GROWTH GAMBLE
bosses and the advice they get from consultants, academics, and cur-
rent received wisdom. In other words, the second reason for failure
stems from the way these managers think about what they are doing.
In this chapter we will focus primarily on the avoidable reasons for
failure. Before that, however, we will first describe the context in which
these managers are working and the inherent risks involved in new
businesses.
THE CORPORATE LIFE CYCLE
The companies we were researching were all mature businesses: a

major oil company, a utility, a consumer products company, a hospital-
ity company, a chemicals company, a restaurant chain, a high-tech
company. The companies had been built many years before by man-
agers no longer involved in them.
These companies had all succeeded in their industries. Based on
superior technology, superior management skills, or a superior busi-
ness model, these companies were all leaders. But, like Intel or
McDonald’s, they were facing a slower-growth future. To maintain
their track record or meet the ambitions of their managers, they
needed to grow or acquire some new businesses.
Intel and McDonald’s were first timers. Their efforts to develop
into new businesses were the first both companies had made. The core
business had been kind to managers, providing plenty of opportunity
over the previous 20 or 40 years. Inevitably, however, managers were
steeped in the management attitudes and processes needed to make a
success of the core and less sensitive to the demands of different busi-
nesses. Some of the other companies we shadowed were second, third,
or fourth timers. Previous attempts had not all been unsuccessful.
Some had faced periods of underperformance and retrenchment.
These managers were inevitably less dominated by one way of doing
business and more sensitive to the demands of different businesses.
Whitbread, for example, had started life as a beer company, and
expanded into pubs as part of its beer strategy. It then began convert-
45
ing its pubs into restaurants and hotels. It also acquired some restau-
rant and hotel businesses, such as Pizza Hut and Marriott in the UK.
In the late 1990s, it sold its beer business and entered the leisure indus-
try through the acquisition of David Lloyd tennis clubs. There had
been plenty of failures along the way and the total value created by this
transformation was not obviously positive. However, some painful les-
sons had been learnt and managers approached the task of creating new
businesses with a good deal of caution.
Whitbread’s experience of new businesses over a period of 30 years
can be contrasted to that of a gas utility company that had been “pri-
vatized” a few years earlier. After privatization the company had
invested aggressively in international exploration, channeling money
from its home-base monopoly to the new area. At the time we were
shadowing the business, management had just consolidated its reason-
ably successful international operations and concluded that additional
growth was still required. Managers had successfully exploited the
obvious growth opportunity and, buoyed with their success, were now
eagerly looking for another growth platform.
The traditional business life cycle of early-stage development, high
growth, maturity, and decline was only partly evident in the companies
studied. Most were past the high-growth stage, at least in the busi-
nesses that had launched the company. Most had some mature busi-
nesses and some growth businesses, but still found the combination
insufficient for their ambitions. Most had already exploited the obvious
adjacent opportunities—restaurants in Whitbread and international
exploration in the gas utility. They were therefore facing a tough chal-
lenge: What should they do next?
INHERENT RISKS
One story from our research illustrates well the combination of chance
events that needs to occur for a new business to succeed. One of the
most successful parts of the Royal Bank of Scotland Group, the
second-largest UK-based bank, is a business called Direct Line. It sells
46
THE GROWTH GAMBLE
motor insurance, motor rescue, and general insurance products direct

to the public and through partner relationships with intermediaries,
such as supermarkets.
In 1980 two managers, Mike Flaherty and Roy Haveland, were
working in the IT department of Alexander Howden, an insurance
brokers. Mike reported to Roy, the head of IT, and Roy reported to
Peter Wood, who was the head of group services. Explained Mike
Flaherty:
At the time there were some important changes in technology happen-

ing—relational databases, a major shift in mainframe technology which
meant that mainframes, that had previously cost £2 million, were now
available second hand for £20,000, changes in drive technology, the
arrival of laser printers and the development of software for composing
documents. This was making it possible to have multiple indexes. If
there was a train strike and you wanted to identify all of your employ-
ees in the affected postal addresses, it would previously have taken days
if not weeks. It could now be done in a day.
In an attempt to justify a large computer, exploit these new technolo-

gies, fulfill the entrepreneurial instincts of the managers involved, and
prove the value of the IT department to a skeptical management, the
department started to sell services outside. This involved manipulating
databases and printing low-volume documents. They did internal cat-
alogs for the BBC and the electoral register for the Borough of
Croydon.
It was probably one of these rare situations where you have a combina-
tion of people with technical and commercial skills. We had all been
detailed systems programmers. We were all more fluent in Assembler
language than we were in English. We could write the machine code
that made our computers sing. But we were also commercial.
In 1982, the team was approached by two people who had worked in
an insurance brokers, selling motor insurance. “They originally asked
47
us if we could print rate books for them a couple of times per year.” As
the relationship developed, the motor insurance brokers, who had also
been in IT, recognized Mike and Roy’s capabilities. The brokers
explained that up until that point the rate books only allowed £5 step
changes. Because of the competitiveness in the industry, it was felt that
it would be a big advantage to be able to quote rates between these
numbers. Their idea was to produce more fine-grained rates and
download them for a fee onto “midi” computers that they would sell to
insurance brokers.
At the time, the brokers’ systems asked 16 questions and gave rates with
step changes of £5. Their idea was to ask 21 questions and give step
changes of 50 pence. We were capable of doing this number crunching
because we had this huge, under-utilised computer.
The business proposition was to sell midi computers and software to

the brokers at cost and then make money by selling them the ratings.
But the cost of midi computers meant that they were not able to pro-
duce a package that was attractive enough for the brokers and the
whole idea fell apart.
The big pay-off from this experience was that Mike and Roy
learnt about producing rate cards. They considered producing rate
cards for Alexander Howden’s motor insurance department, but at
the time Alexander Howden had just been acquired by Alexander and
Alexander. The acquisition had unearthed some funny accounting
and the whole company was in turmoil. Costs were being cut and
Peter Wood, along with other administrative staff, was worried about
his job. As a result, Peter suggested that Mike, Roy, and he try setting
up an independent motor insurance service using this new
technology.
Peter and Roy knew some ex-Howden individuals who had set up as
independent brokers. Working with them, a business plan was devel-
oped. The plan was to set up as a direct insurer to avoid the commis-
sions to brokers of 20–30%.
48
THE GROWTH GAMBLE
We had seen that people were beginning to book theatres and things
over the telephone with credit cards. We just thought of it as a process-
ing task. We could have been just a panel on the AA platform. But we
decided to set up a whole business, with the underwriters as suppliers.
The team spent 18 months looking for a backer for this idea, while
they were still working for Howden. Then they got a break. One of the
jobs they had carried out earlier was data entry of Time Life invoices.
The person they had been dealing with at Time Life became finance
director at the Royal Bank of Scotland (RBS). Peter Wood was talking
to this person about doing printing work for RBS and one day, when
he was playing golf with him, Peter said that he had this idea for a new
business.
The Royal Bank decided to back the idea and, 20 years later, the
business is worth around £2 billion. As Mike Flaherty reflected:
If you had 40,000 people in the UK who knew about relational data-
bases and motor insurance, the chances of one of them coming up with
what turned out to be Direct Line is about as close to zero as you can
get. The number of chance events that had to come together to get this
to happen was mind numbing.
In particular, chance had a big impact on the combination of people

and experience necessary for this business to work. The excess com-
puter capacity in the hands of entrepreneurially minded IT experts
who were under pressure to demonstrate their value. The chance
encounter with the insurance brokers who could see the opportunity to
produce fine-grained rate cards. The commitment of Peter Wood to
this unpromising project because of internal cost cutting.
In addition, Mike Flaherty commented about the importance of the
balance of skills and power among the three founders:
We were a unique combination of people with the right balance of

skills. Roy and I had agreed that we would share whatever we got out
of it. As a result we were the ideal counter balance to Peter.
49
Without the deep IT knowledge, the motor insurance knowledge, the

“seniority” that Wood brought to the team, and the balance between the
team members, the business would never have succeeded. In fact, even
after the Royal Bank’s investment, it had a difficult first three years, only
avoiding closure because of Wood’s negotiation and persuasion skills.
The Direct Line story is important because it can be contrasted with
those of the 10 or 20 other UK financial institutions that tried to build
large direct businesses at about the same time. Spotting the opportunity
evident from Direct Line’s early successes, these companies set up proj-
ects, ventures, and divisions aimed at the direct market. But as Mike
Flaherty explained, the chances of any of these companies assembling
the unique skills and experiences that fueled Direct Line were zero.
Inevitably, there were other paths into this market and some succeeded.
However, none has come close to the success of Direct Line.
Not all of our examples were dominated by so many chance events.
Some, like IBM’s decision to get into the PC business, were businesses
waiting to happen, and were previewed by a number of earlier
attempts. Others, like the decision by the electrical retailer Dixons to
set up PC World, were entrepreneurially opportunistic. The CEO
noted the success of a new start-up, acquired the business, and built on
its business model.
However, anyone who has studied new businesses is struck by the
large number of different things that need to come together to make
the business a success. Even if the probability of each element coming
good is 80%, the overall probability of five elements coming good is
32%, and 10 elements is 11%. In other words, new business ventures
are inherently risky.
AVOIDABLE CAUSES OF FAILURE
In addition to the inherent risks of any initiative to enter new busi-

nesses, we have observed some common and avoidable problems that
lead to value destruction. They are problems that stem from the way
managers think about the task (Figure 3.1).
50
THE GROWTH GAMBLE
Figure 3.1 Avoidable causes of failure
THE ICARUS PITFALL
Icarus, according to Greek myth, was trying to escape from the island
were he was held captive. He and his father glued feathers onto their
arms and used them to fly. However, Icarus became overambitious and
tried to fly too high. The heat of the sun melted the glue holding his
feathers, and he fell to the ground. The efforts by some of the man-
agers we shadowed felt rather Icarus-like.
There are a number of reasons managers try to fly too high. The
first comes from the way strategic plans are developed. Managers,
developing a strategy for a company like Intel or McDonald’s, will start
with the plans of the major businesses. These are aggregated into a
corporate plan and an assessment is made about whether the company
has sufficient capital to support the plans of the businesses and whether
the performance that will result is good enough.
Good enough for an ambitious management team normally means
top-quartile performance among a peer group of similar companies.
Double-digit annual growth in earnings is not an unusual ambition.
When the existing businesses cannot deliver this desired growth level,
managers consider that the gap needs to be made up with new busi-
nesses. Hence, companies often derive their ambition for new busi-
51
nesses from an analysis of the gap between their likely performance

and their ambition. This is like a traveler in the desert deciding how
much he would like to drink based on his level of thirst rather than the
amount of water he has available. The alternative approach, the one we
recommend, is to develop the ambition from an assessment of the
opportunities.
BG plc was a classic example of this. In the late 1990s the company
consisted of two businesses: a UK domestic gas-transmission business
that owned pipelines and storage facilities and an international gas-
exploration business that owned gas reserves. Both businesses required
focused attention. The domestic business was regulated and manage-
ment needed to focus on driving costs down faster than the regulator
had predicted. The international business was created out of a merger
between a pure exploration business and a downstream business
involved in gas terminals, pipelines, and power stations. The inter-
national business needed to focus on making this merger work.
The holding company managers, including the previous head of the
international distribution business, were reducing central costs and
decentralizing functions and decisions to these two divisions. However,
when the central team produced a corporate plan, they concluded that
the combined performance of the two divisions would fall well short of
their ambitions.
As a result, they set up a corporate development division, incorpo-
rating a property development activity, a central research function,
some shared services, and responsibility for new businesses. The head
of the division, Stephen Brandon, commented:
We have set ourselves a very big ambitious target. We want to be in the

top quartile of the FTSE 100 in performance. Based on the CFO’s
analysis this means that we need to create an additional £500 million of
net income or the equivalent in value within five years.
I have to find businesses that create value at a phenomenal speed.

There are only two ways. Either major acquisitions, like the BP/Amoco
deal, or something new in an area that can deliver huge increases in
52
THE GROWTH GAMBLE
value. In the current environment (1999), this probably means the

Internet or telecoms.
Moreover, the ambition was not just hope. The bonuses of the top
managers were linked to the “top-quartile” goal.
The development division screened some 24 new business ideas and
decided to take 11 to the next stage. One, investing in the telecoms
industry, was the most attractive, in part because it had the most poten-
tial of achieving the ambition. Within a year, the 11 ideas had been
presented to a management conference in the form of a “growth fair.”
All but one of the ideas attracted a sponsor, the telecoms project being
the most significant. Two years later, part of the telecoms investment
had been closed or sold at a cost of £350 million, and none of the other
new business ideas had made significant progress.4
Undoubtedly the size of the ambition influenced the thinking and
efforts of the development team, but not in a beneficial way. It caused
them to look at other companies, such as Vivendi and Enron, which had
created new value quickly. It encouraged them to consider industries
where BG had little experience. It also encouraged them to do things
quickly, reducing the amount of time they had to test and learn. They tried
to fly too high because of a focus on the gap rather than the opportunity.
The second reason for flying too high is pressure from the financial
community. The BG story is an example of the pressure that is applied
both directly and indirectly by the financial community. At BG, the top
team were clearly influenced by the performance that they thought was
expected of them by “the city.”
Unfortunately, at certain periods the financial markets are influ-
enced by fashion. When some companies are growing by acquisition,
all companies are asked what their acquisition plans are. When some
companies are setting up e-commerce incubators, all companies are
asked to describe their e-commerce plans. The combination of com-
parative target setting and the pressure of fashion can easily combine
to encourage managers to set goals that are overambitious.
The third motivation to fly too high is the belief in the power of
stretch goals already discussed in the previous chapter. Unhappily, as
53
with the BG case, a stretch goal is less than useful. When managers are
doing something they are familiar with, a big goal will help raise their
heads and encourage them to consider changes to their routine. But
when managers are doing something they have not done before, they
do not need to be jolted out of a routine, they need to be cautious.
The fourth reason for flying too high is a concern for long-term
survival. We will address this more fully in the next chapter, but it is
well captured by the quote from McDonald’s: “the misery of uncer-
tainty is less than the certainty of misery.” Managers believe that they
have an obligation to find new businesses for their companies almost
regardless of the risks. Without these new businesses, they worry that
their company will mature and ultimately become obsolete. This fear
can override the caution that normally controls management action.
We believe that companies should resist these pressures. Instead of
setting high ambitions for new businesses, they should set targets only
after they have reviewed the opportunities. There is something flat
footed about using gap analysis or survival pressures to set targets. It
implies that managers have a goal that is independent of their environ-
ment. It suggests that a company, with many opportunities, can stop
investing in them once the goal has been met. It also suggests that
companies with few opportunities can conjure up new ones until the
goal is met. It is much more sensible, we believe, to set targets only
after understanding what the opportunities are.
THE HELEN OF TROY PITFALL
Helen, according to myth, was the beautiful wife of the king of

Macedonia. She fell for the prince of another country and went with
him to his home city, Troy. Her husband assembled an army and
besieged Troy for some years in an attempt to get her back. He only
finally succeeded with the deception of a hollow wooden horse in
which he concealed soldiers. He pretended to abandon the siege, offer-
ing the horse as a sign of his good intentions. When it was brought
inside the gates of Troy, the concealed soldiers sneaked out of the horse
and opened the city’s gates to the attacking army.
54
THE GROWTH GAMBLE
The myth is probably more famous for the Trojan horse than for
Helen. However, it was the pursuit of her beauty that caused all the
damage in the first place. In a similar way, we have seen managers,
obsessed by the attractions of a market opportunity, become blinded to
its risks or its appropriateness.
The Helen of Troy pitfall stems from a combination of the seduc-
tiveness of “sexy” markets and a failure to apply the available tools of
objective analysis. When analyzing new business opportunities, we have
noted that managers devote a lot of energy to searching for large, grow-
ing, and profitable sectors. By definition these are relatively rare, so that
when one is identified it has a magnetic pull that is hard to resist.
Take the example of British Sugar, one of two suppliers of sugar in
the regulated UK market. Faced with a mature market and possible
changes in the regulated structure, the company (a division of ABF)
was actively seeking new business ideas. The head of the development
unit was looking at opportunities with some connection to the core
skills of the business: the processing of agricultural raw materials.
Unfortunately, there are few large, profitable, growing sectors in pri-
mary agricultural products.
However, the team identified that automotive companies faced leg-
islation requiring them to produce vehicles with a greater percentage
of recyclable or biodegradable parts. One area where biodegradability
was being considered was the interiors, using natural fibers such as
hemp. Armed with the prospect of a fast-growing market, the team
looked closely at the industry and identified one of the leading hemp
companies as a prospect. This company was in a poor financial condi-
tion and was looking for a partner.
Excited by the growth prospects, the team pitched the idea to sen-
ior management on more than one occasion. They abandoned it only
when an external consultant reported on the poor prospects for the
company and the sector. In their enthusiasm, the team had overlooked
evidence suggesting that the hemp industry traditionally had low prof-
itability, that this growing segment would be selling into an industry
with the toughest buying policies in the world, and that there was lit-
tle technical connection between hemp and sugar.
55
Another force driving the hemp proposal was the feeling that envi-
ronmentally friendly solutions are part of the future. British Sugar, the
team implied, should be investing in this kind of project because it
would provide more long-term security for the company. Similar
thoughts influenced many of the other teams we were shadowing. “We
should be moving into services.” “We need to be in China.” “We must
be in the faster-growing sectors.” “Something to do with the internet.”
All these are thoughts built round a belief that companies should invest
in the next big trend or even catch onto the current trend.
The awkward part of this thinking is that it is hard to dismiss. “You
mean we should be investing in yesterday’s ideas?” “Are you suggesting
that we do not try to look into the future?” These are common
responses to a suggestion that chasing the next wave is unlikely to
deliver the desired result.
The next wave, we argue, will attract more interest and investment
than other businesses. Hence, all things being equal, the next wave is
likely to be less attractive than other opportunities, unless certain con-
ditions exist: unless the company has some special advantage over
rivals; unless there are exceptional benefits to be gained from being an
early investor; or unless the opportunity is one of the rare situations
where demand is likely to exceed supply for a few years. Hence, unless
these conditions exist, chasing the next wave is likely to destroy value,
not create it.
The second reason for the Helen of Troy pitfall is management’s
failure to use appropriate tools of analysis. Too much attention is given
to growth rates and business model economics (whether the service
costs less than the customer is prepared to pay), and not enough atten-
tion is given to what Michael Porter calls the Five Forces.5
We shall say more about the Five Forces in Chapter 5. However,
the main point of Five Forces analysis is to predict the future profit lev-
els in an industry sector. The prediction depends on judgments about
five forces: bargaining power of customers, bargaining power of sup-
pliers, threat of new entrants, threat of substitutes, and intensity of
rivalry between competitors. Each of these forces can be favorable to
or damaging to high profitability. Companies will find it harder to
56
THE GROWTH GAMBLE
make profits in sectors where one or more of the forces is against

profitability.
The power of Porter’s tool is that it excludes the one measure that
so easily seduces the inexperienced—growth. The tool helps pinpoint
declining sectors that are likely to provide more profit potential than
growing sectors. It also helps managers distinguish between those
growing sectors that are likely to reward companies that invest, and
those that are likely to punish companies.
Despite the tool being widely taught and included in every textbook
on strategy, it was not formally used by any of the teams we shadowed.
This does not mean it was completely ignored. Issues such as the
power of customers and the likelihood of competitor response were
often discussed. But in our view, because the tool was not used for-
mally, the seductiveness of a growth sector and of businesses of the
future frequently crowded out Five Forces thinking. Without such a
tool, it is hard for managers to resist the temptations of the Helen of
Troy pitfall.
HUBRIS
Hubris is the inappropriate self-confidence that can develop in suc-

cessful management teams. The Concise Oxford Dictionary defines
hubris as “insolent pride or security” or “overweening pride that leads
to nemesis.” Nemesis was the goddess of retribution. She made sure
that those whose pride caused them to challenge the gods suffered.
Unfortunately, the outcome of many new businesses is so disap-
pointing that one could be forgiven for thinking that Nemesis was tak-
ing her revenge on the insolence that caused the managers to believe
they could beat the odds.
One manufacturing company had two such experiences. The com-
pany was involved in a processing activity that generated excess heat
and excess carbon dioxide. As managers looked for ways to use this
excess, they came into contact with the tomato industry.
In temperate climates, tomatoes are often grown in heated
glasshouses. Moreover, they grow better in a carbon dioxide-rich envi-
57
ronment. The company therefore tried to link up with a tomato

grower to sell its spare heat and carbon dioxide. However, the tomato
growers were not very profitable and had little cash for investing in
new capacity. Attempts to persuade a grower to build glasshouses near
the company’s factory failed. This spawned the idea of using the man-
ufacturing company’s strong cash position to finance the glasshouses
and, at least in part, enter the tomato-growing industry.
The company’s management was, however, very risk averse.
Knowing that it had little experience in tomatoes, the board insisted on
a plan that would reduce risk to a minimum. This involved signing a
contract with one of the major growers to sell all the tomatoes at an
agreed profit margin. This was sufficient to provide a respectable
return on the investment in building the glasshouses.
Inevitably, there were start-up problems. But after a year the project
seemed to be going well. However, unknown to the company, there were
big changes going on in the tomato industry. Supermarkets were chang-
ing to a new variety of tomato with a better shelf life. Customers were
responding favorably to these new tomatoes—ones that could be grown
successfully only in hotter climates. In the second year, there was such a
glut of the old variety that prices only just covered transport costs.
Moreover, it seemed likely that the market would deteriorate further.
At the time of writing the project is still in operation, but its future
is in doubt, especially when the contract with the grower expires.
Looking back, management had considered some negative scenarios,
but none that involved prices that left no margin over transport costs.
About the same time, the company decided to invest in a combined
heat and power plant. This would save energy costs in its processing
operation and generate electricity that could be sold into the local elec-
tricity grid. The project was low risk. The technology was mature. The
cost savings were known. There were buyers for the electricity happy
to sign long-term contracts. Furthermore, the initiative provided the
opportunity for the company to experiment in the power industry, an
area that had been considered as a diversification.
In the final approval process, some senior board members argued
against the long-term electricity contracts. They believed that it was
58
THE GROWTH GAMBLE
unnecessary to eliminate all risk. They argued that the contracts were
overly favorable to the buyers given the stability of the electricity mar-
ket. The project therefore went ahead with only the minimum of long-
term contracts.
The market, however, did not remain stable. As the company was
building the unit, other companies were building similar units and
electricity companies were constructing gas-powered generators of
their own. The result was a glut of electricity capacity.
As with the tomato experiment, the price of electricity fell precipi-
tously. For a period, it was below the cost of the gas needed to gener-
ate it, and the project was losing money even before deducting the cost
of capital.
With two such improbable disasters on their hands, the manage-
ment of the company could well conclude that Nemesis was taking
revenge on them for attempting to get into different businesses.
The primary cause of hubris is the lack of a framework for deciding
whether the company has the appropriate capabilities, experience, and
knowledge for the new business. The problem stems from focusing
primarily on areas of fit rather than taking a balanced view of a com-
pany’s full set of skills.
There are three common tools for assessing fit. The first views fit
in terms of technology or market. The product/market matrix6 (Figure
3.2 overleaf) has market and product (or technology) as its two dimen-
sions. Development teams using the matrix ask “What other products
can we sell to our current customers?” or “What other markets can we
enter with our existing technology?” If a new initiative only involves
changing one of these two dimensions, it is considered to offer suffi-
cient fit.
A second approach involves defining a company’s core competen-
cies.7 These may be broad functional skills, for example Honda’s skill
at small-horsepower engines or McDonald’s skill at managing its sup-
ply chain. The skills may also be narrow technical competencies, such
as 3M’s thin film coating technology, or special market relationships,
such as Shell’s relationships with certain governments. If the new ini-
tiative involves one of these core competencies, it is considered to fit.
59
Figure 3.2 The product/market matrix
A third tool of analysis is the “adjacencies steps test.”8 This defines

six dimensions of adjacency: customer, competitor, channel, geo-
graphy, cost structure, core capability. When any dimension is differ-
ent from the company’s core business, it is recorded as one step away.
If all six are different, the new business is six steps away and the prob-
ability of success is miniscule. The more steps, the lower the probabil-
ity of success (Figure 3.3).
The first two approaches to fit analysis suffer from giving too much
attention to what fits and not enough to what does not fit. In contrast,
the adjacencies steps test focuses only on what does not fit. In our view
managers need a balanced understanding of what fits and what does
not fit.
There are two levels in the organization where fit issues need to be
analyzed: the business unit level and the parent company level. The key
questions managers need to ask are:
1 Does our company have some contribution to make to this new

business that is more important to success than the contributions of
current or likely future competitors?
2 Will the managers we put in charge of this new business, taking
account of the habits, instincts, and experience that they already
60
THE GROWTH GAMBLE
Figure 3.3 Adjacency steps
have, understand the business as well and be as capable of making a

success as their current and likely future competitors?
3 Will the managers this business reports to, taking account of their
habits, instincts, and experience, understand the new business and
be as capable of helping the new business as the parent companies,
venture capitalists, or other financiers of competitor businesses?
Time and again we were told stories of misjudgments and mistakes

caused by the inexperience of business-level managers. Mars’s move
into ice cream in Europe is an example. After a successful test market,
Mars followed its normal strategy of producing easy-to-wrap bars.
This fit the company’s manufacturing skills and its desire to be the low-
cost producer. Concerns were raised about whether consumers would
accept ice-cream bars without sticks. But evidence from the UK sug-
gested that they could be persuaded.
As it turned out, consumers, especially in France and Germany, did
prefer sticks in their ice creams and were slow to try the new products.
61
Without long experience of consumer preferences in this market, it

was easy for Mars’s managers to misjudge the reaction.
We have also been regaled by complaints that parent company man-
agers do not understand a new business. Inappropriate policies and mis-
directed guidance are imposed from above in a way that makes it hard
for the business to succeed. For example, when the tobacco company
BAT entered the financial services industry, parent company managers
believed they could apply the marketing skills of the tobacco industry in
financial services. One of the core concepts in tobacco marketing is rel-
ative market share. Hence BAT encouraged its newly acquired financial
service businesses to analyze relative market share. This exposed poten-
tial for increasing share and, over the next three years, the new busi-
nesses aggressively gained share. However, in the fourth year the
businesses announced huge insurance losses, and a few years later BAT
exited financial services and refocused on its tobacco interests.
Unfortunately, in financial services market share is not always prof-
itable. In its aggressive drive for share, BAT had inadvertently encour-
aged its new businesses to take on high-risk customers. When the
management team of the financial services business was asked why they
had agreed to this policy, when they knew it involved highly risky
transactions, they replied that they had been concerned, but felt they
ought to give their new parent’s policies a try. “We thought they prob-
ably knew something we did not. After all, they had bought us.”
We will explain in Chapter 5 how we believe managers can reduce
the risk of hubris. However, it is hard to eliminate altogether.
Whatever framework or analysis structure is used, it can be polluted by
excessive optimism. The good news is that a cautious mindset and
careful use of screening methods can greatly reduce the tendency for
managers to misjudge their capabilities.
THE NUMBERS GAME
The numbers game is the final cause of avoidable error. We mentioned

this in Chapter 2. Many managers believe they will succeed only if they
have many attempts. This encourages them to devote considerable
62
THE GROWTH GAMBLE
resources to generating and nurturing ideas, running incubators and

venture capital units, and supporting a portfolio of new ventures.
This approach to developing new businesses is encouraged by most
textbooks and partly supported by Gary Hamel in his Foreword to this
book.9 The argument is compelling. Progress comes from a process of
variety followed by selection. If a company is to progress, it needs to
generate sufficient variety to give it a chance of finding an exciting new
business. Once this has been found, it can then select and commit to
this new opportunity. Because only a small percentage of new business
ideas turn out to be substantial opportunities, the company needs to
invest in a large number to be confident of finding at least one.
The venture capital industry, it is argued, is proof of the need to
play the numbers game. Venture capital companies invest in a large
number of projects before they come up with one real winner.
Companies, it is claimed, need to recreate Silicon Valley (the place
where venture capital is most vibrant) within their organizations.10
Unfortunately, despite the power of the logic, the numbers game
does not appear to work. None of the companies we shadowed was suc-
cessful with the approach of launching a large number of ventures.
This was a surprise to us. As a result, we extended the research to
include a sample of corporate incubators and venturing units. We even
combined our research with work that was happening in parallel at the
London Business School.11 This gave us a sample of over 100 corporate
incubators and venturing units: vehicles set up by their parent compa-
nies to promote, nurture, and invest in new ventures.
The vast majority of these units were unsuccessful, even in achiev-
ing the limited objective of earning a return on the money invested.
Less than 5% of them achieved the more demanding objective of suc-
cessfully spawning one or more significant new businesses for their
parent company. Moreover, there is little disagreement about these
results. Research by Bain & Co. and by Strategos reached similar
conclusions.12
The common response to this dismal record is to accuse companies
of failing to implement the venture capital model effectively. If the
venture capitalists in Silicon Valley can succeed, runs the argument,
63
then companies should be able to do as well so long as they effectively

copy the Silicon Valley model.
In our view this is the wrong conclusion to draw. We will explain
more about why the incubation/venturing approach does not work in
Chapter 8 and we will propose a different approach based on rigorous
selection in Chapter 5. We believe that the only way to beat the odds
is to reject the numbers game and develop a much more demanding
selection process.
WHY DO NEW BUSINESSES FAIL?
One of the objectives of our research was to find out why developing
significant new businesses is so difficult. In previous work we had done,
we had observed that companies with a spread of businesses were only
successful when the skills and resources of the parent company fit well
with the needs of the businesses (see Chapter 6 for more on parenting
theory). The simple logic of parenting theory directs managers to
enter new businesses if they have appropriate parenting skills and to
avoid businesses where the parenting skills do not fit. We believed that
this logic for being in different businesses is quite widely understood.
Hence we were somewhat puzzled by the high failure rates of man-
agers tasked with extending their portfolios.
This chapter goes some way to explain why managers have not fol-
lowed the simple rules of parenting theory. The pressures on them,
their managerial instincts, the advice they are getting from other quar-
ters, the role models they have, and the approach to the task that seems
appropriate are all part of the problem. To improve, managers need to
be given guidance they can believe in and a tool or a theory that is built
on the basic factors involved in business success. The following chap-
ters are an attempt to provide the necessary frameworks and guiding
thoughts.
64
CHAPTER 4
WHEN LOW GROWTH IS

BETTER THAN GAMBLING
T
he story so far is that most companies fail when they try to
enter new businesses. The main problem is that there are few
significant opportunities that fit the specific set of resources
each company possesses. A secondary problem is that managers are
driven by some inappropriate motivations and beliefs. Our conclusion
is that managers should be more selective rather than more energetic
in their approach to new businesses. They should assess opportunities
to enter new businesses with a tough set of criteria, only investing in
those that pass a strategic business screen.
The implication of this advice is that some companies will, at some
points in time, screen out all ideas for new businesses. This will con-
demn the company to a rate of growth determined by its core business.
If the core market is growing slowly, the company is likely to be low
growth, at least until the screening process identifies some promising
opportunity. If the core market is in decline, the company is likely to
be in decline.
This chapter will address this low-growth challenge. It will not offer
solutions that will help a company become high growth. Rather, it will
address the issue of whether low growth is such a bad thing, and help
managers decide when they should choose low growth rather than
gamble on new businesses.
Most managers believe growth to be an imperative. It feels natural
to have an ambition to grow. Growth is like a virtue, something of
unquestioned value. But growth is also a dilemma. As Clayton
Christensen, Harvard’s leading business strategist, points out:
WHEN LOW GROWTH IS BETTER THAN GAMBLING
Most executives [in low-growth businesses] are in a no-win situation:

equity markets demand that they grow, but it’s hard to know how to
grow.
In this chapter we are going to challenge the managerial presumption

that growth is necessary. We are going to argue that, in some circum-
stances, it is acceptable to choose to be a low-growth company. Before
we present our thinking, however, it is worth first summarizing the
arguments in favor of growth.
THE ARGUMENTS FOR GROWTH
There are two sets of arguments for growth: value driven and moral.
The value-driven arguments are powerful. First, the value of a com-
pany, in terms of market capitalization, is greatly affected by its growth
rate. For companies with modest growth rates, each percentage point
of growth will add around 10% to market capitalization.1 If a decision
not to grow reduces expected future growth rates, the value of the
company will decline, sometimes dramatically. For example, the aver-
age market capitalization decline for companies that hit a “growth
stall” is over 50%.2
Second, success breeds success. A decision to stop growing may
break the magic of momentum, allowing managers and employees to
lower their sights and encouraging them to be satisfied with less. Sir
Clive Thomson, CEO of Rentokil Initial for 20 years and nicknamed
“Mr 20%” because of his growth record, explained:
Managers began to believe that this was the norm, that they could do
it every year, and they achieved much more as a result.
Third, growing companies attract the best talent because they can pro-
vide more interesting career opportunities. Once growth slows the
opportunities for advancement decline, and capable younger managers
are likely to look elsewhere. A manager at Kerry Foods described her
66
THE GROWTH GAMBLE
feelings when the company reduced its growth target from 15% to
10%:
I felt cheated. I felt that we were no longer ambitious to be a top com-

pany. I was proud of our 15% record. Now I do not know what to be
proud of.
The moral arguments are often deeply held, but strike us as less power-
ful. The assertion is that, if everyone stopped investing in new busi-
nesses, progress would slow down. Therefore, companies have a duty
to society to innovate and develop new businesses. Moreover, societal
benefit can result even if the efforts of a company do not create share-
holder value. As economists point out, all commercial activities create
a consumer surplus (the value to consumers that exceeds the cost of the
product) as a well as a producer surplus (the net profit that the pro-
ducers make after deducting the cost of capital). It is possible to create
a surplus for society even when the producer surplus is negative.
Hence, the moral argument is that companies should innovate even if
the producer surplus is questionable.
It may seem implausible that managers might make a moral argu-
ment in favor of new businesses, but it is not. At least half of the man-
agers we shadowed used the moral argument in one form or another at
some point in our relationship. With hindsight, this is not surprising.
Managers who get selected to promote new businesses are those who
believe it is important and who get turned on by creating new things.
One head of new businesses described the excitement she felt when
driving up to one of her new ventures and compared it with driving up
to the company’s headquarters. The new venture was on a bumpy road.
It led to some buildings in poor repair. Inside, managers were working
alongside front-line employees with a passion for their product.
Managers have values and emotions just like the rest of us, and those
who are excited by new businesses are as comfortable making a moral
connection with their work as anybody else.
One manager drew an analogy with the efficient market hypothesis.
Fewer companies investing in new businesses would, he argued, reduce
67
the volume of new businesses being launched, which would reduce the
“liquidity” of the new businesses market.
Another manager argued that companies are living organisms that
have value separate from their economic role. They are part of the rich
tapestry of diversity and are communities that provide satisfaction for
their members. It may be appropriate, therefore, for companies to
make value-destroying investments in pursuit of survival.
Another manager, an ex-CEO of a major European company, was
concerned about competition from Asia. He argued that Asia is more
committed to new businesses and more comfortable betting against the
odds. He would like to see similar determination being exhibited by
European companies:
If we do not match their appetite for risk, we will create an economy

with a permanent underclass with no prospect of getting jobs.
These moral arguments are inadvertently supported by most authors.

Explicitly or implicitly, they presume that companies should seek to
survive over the long term. Since all businesses have finite life
expectancies, the only sure way to survive is to migrate to new busi-
nesses with better prospects before the existing ones die. In fact,
research into companies that have survived for more than 100 years
demonstrates this. They all radically changed their portfolio of busi-
nesses at one or more points in their history.3
This recognition of the fragility of existing businesses suggests
that companies must develop the skills needed to get into new busi-
nesses. It is not a question of whether to get into new businesses, but
how to get into new businesses. As Hein Schreuder, corporate vice-
president, corporate planning for Dutch chemicals company DSM
complained:
If you argue that our company should stick with its existing businesses
and cut back on its new businesses, you are asking us to commit pre-
mature suicide.
68
THE GROWTH GAMBLE
Mats Lederhausen of McDonald’s expressed the same thought when he

declared that “the misery of uncertainty is less than the certainty of
misery.”
In this section we will address these moral arguments. In the rest of
the chapter, we will tackle the value-driven arguments.
The moral arguments are deeply felt and cannot be rejected out of
hand. But they are, we believe, based on a presumption that better
selection is not possible, that managers must be driven by hope rather
than by strategic logic.
We counter the moral argument in three ways. First, initiatives to
create new businesses have societal costs as well as benefits. A new
business that fails creates problems for society in the same way that a
failing old business does. Hence we are looking at the balance of cost
and benefit. A company that starts many new businesses, most of which
fail or underperform, is not benefiting society even if it survives a few
years longer than it would otherwise have.
Second, given the odds of creating new businesses, a strategy of
careful selection between alternative new business projects is likely to
create more value than one of ambitious expansion. This is not only
more value for the company, helping it to survive, but more value for
society, both in reducing the number of failures and in helping apply
resources to their most productive uses. Companies that try to get into
new businesses indiscriminately are gamblers squandering society’s
resources. They are as immoral as those who have resources and do not
use them. If better selection could lead to fewer failures and more suc-
cesses, we would all be happy. So the question comes down to an issue
of whether it is possible to eliminate some of the losing projects before
they are started.
Third, the existing managers of an existing business are not neces-
sarily the managers best able to exploit some new opportunity, even
though it may draw on links with their existing business. For example,
in flat-panel displays Philips had a technical lead. However, it also had
a location and managerial disadvantage. The cluster of capabilities
needed to support the business was most available in Asia, not Europe,
giving Japanese and Korean competitors an advantage. Moreover,
69
Philips did not have a team of managers with all the skills needed to
champion the flat-panel business. Faced with a location disadvantage
and only adequate management, Philips struggled. After a few years, it
merged its flat-panel business with LG. In this situation, investing in
the flat-panel business in Europe, even with commitment, was the
equivalent of chasing rainbows.
Instead of encouraging managers to deploy resources to enter new
businesses, society might benefit more by encouraging them to select
between those battles they can win and those where they will do better
by selling existing resources to a third party. In our view, those who do
not know how to turn their spare resources into successful activities
have a responsibility to give the resources back to the market so that
others who do know what to do can use them. This is unlikely to hap-
pen unless we have a rigorous way of screening opportunities and a
philosophy of growth that makes it acceptable, in certain circum-
stances, to choose not to grow.
Clearly, there are non-economic benefits for the continuing exis-
tence of some business organizations. If so, it is better that these organ-
izations are funded by the communities gaining the benefits. There are
many governance structures, such as partnerships, trusts, charities, and
cooperatives, that can legitimately give credit for non-economic bene-
fits. Companies funded by capital markets are not well placed to take
on this role. They are best advised to act rationally, choosing to invest
when the probability of success is high and choosing to avoid new busi-
nesses when the odds are stacked against them.
HOW COMMON IS LOW GROWTH?
Turning to the value-driven arguments, the rest of this chapter will

explain that a decision to reduce growth or even not to grow at all may
be the right choice. We show that low-growth companies and even
negative-growth companies can still produce acceptable financial
results. Low growth, we argue, is the norm, not the exception.
Managers need to recognize that many companies at many points in
70
THE GROWTH GAMBLE
their history will be able to grow only slowly, if at all. Managing low
growth wisely may be a skill that most managers still have to learn.
The world economy grows in real terms at less than 5% per year
and, even if the commercial sector is growing faster than the govern-
ment sector, the average real growth of the commercial sector in most
economies is below 5% per year. Even high-growth economies, such as
China, do not sustain growth rates of more than 7–8% for many years.
In Europe, real growth in the last decade has been less than 3% for
many countries. With underlying economic activity growing at these
sorts of rates, most companies must perform similarly. In some
economies and some industries inflation can come to the rescue.
Companies can grow revenues at above 10% even when their under-
lying volumes are growing at less than 5%. However, inflation only
masks the truth. It may make it easier for managers to accept low real
growth rates. But it does not change the fact that low real growth is the
norm.
There is, of course, huge variation. Intel and Microsoft sustained
growth rates of over 30% per annum for much of the 1990s, even in an
industry where prices were declining. At the other end of the scale
major industrial companies, like Britain’s chemical group ICI, suffered
decline, with gross revenues falling by 36% in money terms, let alone
real terms, between 1994 and 2003. The one-time US industrial giant
USX performed similarly, with revenues falling by 32% from 1990 to
2000. Between these extremes, stalwarts of the world economy, like
Ford, Procter & Gamble, or Nestlé, have grown on average at less than
5% per annum in real terms, and have not always kept pace with GDP
growth. Between 1989 and 2003, Ford clocked up an average of 2%
per annum real growth in total sales, when the US economy grew by
an average of 2.84% per annum in real terms.
The Corporate Strategy Board’s research on growth stalls is instruc-
tive in this regard.4 It showed that 95% of companies that were at some
point part of the Fortune 50 hit a growth stall. Following the stall, 83%
of companies had growth over the following 10 years of less than 2%.
In other words, nearly 80% of companies that enter the Fortune 50
have faced periods of 10 years when their average growth was less than
71
2%. Moreover, nearly 50% of companies experienced long periods of

negative growth. In an age where growth is everything, this may seem
incredible. Most of the years covered by the analysis, the last half of the
twentieth century, were good years for business. Yet half of the most
successful companies in America experienced long periods of negative
growth. Only 8 companies out of a sample of 161 avoided a growth
stall. Only six companies (Chase Manhattan, Coca-Cola, Fleming,
Johnson & Johnson, Merck, and Motorola) experienced a growth stall
and were able to recover to average growth rates of more than 6%. In
other words—at least for America’s most successful large companies,
the companies in the Fortune 50—high growth is not the norm. Low
growth and real decline in revenues are not uncommon.
The Corporate Strategy Board also documented the growth rates
(deflated by the growth in the economy) of companies before and after
entering the Fortune 50. In the years immediately prior to entry,
growth rates between 9% and 15% faster than the economy were not
unusual. But the average growth rate 2 years after entry was 2% faster
than the economy, 5 years after entry minus 1%, 10 years after entry
plus 3%, and 15 years after entry minus 4%. If you exclude the year of
entry, the highest average growth rate achieved was only 5% faster
than the economy (Figure 4.1 overleaf). Companies do have long peri-
ods of high growth; both Intel and McDonald’s are examples of this.
However, once the core businesses start to slow down, the norm is for
the corporate growth rate to slow despite efforts by managers to find
new growth businesses.
Maybe the Corporate Strategy Board’s research does little more
than document one stage in a corporation’s life cycle: the transition
from high growth to low growth. It appears to be a painful stage and
one that many managers do not seem good at managing: 28% of com-
panies lost 75% or more of their value following a stall, and only 5%
lost less than 25% of their value. This stage, the one when managers
are looking for new growth businesses because the core businesses are
maturing, is the focus of this book. This chapter is therefore designed
to help managers prepare for periods of low growth and manage them
well.
72
THE GROWTH GAMBLE
Figure 4.1 Fortune 50 growth rates
THE FINANCIAL CONSEQUENCES OF LOW

GROWTH
The reasons for growth given at the beginning of this chapter—fall in

stock price, loss of momentum, and loss of talent—are presumed to be
very detrimental to financial performance. Most managers assume that
low-growth companies are failing companies that cannot generate
decent returns for shareholders. Moreover, the catastrophic declines in
market capitalization of companies that hit a growth stall serve to rein-
force the message that low growth is bad. However, this fear of low
growth is a misunderstanding of the way the market anticipates per-
formance. The challenge of producing decent returns for shareholders
year on year is no different for low-growth companies than for high-
growth companies, so long as this is what the market expects.
Some simple modeling will help (Box 4.1 overleaf). This compares
two companies with steady growth rates, one at 10% and another at
3%. The stock price at the beginning of the period is calculated to take
account of the different growth rates, the faster-growing company
obviously having a much higher stock price.
73
Box 4.1 Simple modeling of high and low growth
Suppose we have two companies, A and B, and both pay out dividends of $100 million per
annum.A is expected to grow earnings at 10% p.a., but B at only 3% p.a.The respective annual
returns on these two companies (total shareholder return or TSR%) will be the growth in value
of the company over the next year plus the dividends as a percentage of the original market
value of the company. Suppose also that shareholders can expect an average return on their
market investments of 12%. The following calculations show that TSR% will be the same for
both companies, namely 12%.
Company A Company B
Dividends ($m) 100 100
Expected annual earnings growth in coming year (%) 10% 3%
Market capitalization at start of year ($m) VA VB
Expected increase in market cap over year ($m) 10% of VA 3% of VB
So:
Expected total annual return to shareholders is
increase in value plus dividends ($m) 0.10*VA+100 0.03*VB+100
Suppose shareholders can earn average of 12% on
their stock market investments.So the total annual
return required by shareholders will be ($m) 12% of VA 12% of VB
The market will value companies A and B so that the
expected total annual return to shareholders equals
12% of what they would have to invest, i.e., what
they require on their investment 0.02*VA=100 0.09*VB=100
Hence market capitalization ($m) is VA = 5000 VB = 1111
TSR%, or total annual return to shareholders as a 0.10*5000+100 0.03*1111+100

percentage of market value, is thus as % of 5000 as % of 1111
So TSR% is 600/5000 133/1111

So TSR% is the same in both cases = 12% = 12%
NB: The same result would apply for a company in decline.If Company C had an expected decline
in earnings of 5% p.a., the equation for calculating market capitalization (VC) would be:
-0.05*VC + 100 = 0.12*VC or 100 = 0.17*VC or VC = 588
TSR% would then be (100 - 0.05*588)/588 = 12%.
74
THE GROWTH GAMBLE
What Box 4.1 demonstrates is that the slow-growth company,

assuming all else is equal, produces a total return for shareholders
identical to that of the high-growth company. The return is the same
even with negative growth. This may seem counter-intuitive.
However, the explanation lies in the different starting valuations.
These differences exist because the market anticipates performance.
Companies can only produce a superior total return for shareholders if
they outperform shareholder expectations. If they perform according
to shareholder expectations, they will deliver an average total return. If
they perform less well than expected, the return will be below average.
In other words, the problem of low growth is less about the annual per-
formance challenge and more about major shifts in market percep-
tions. If a company suddenly moves from high growth to low growth,
perceptions will change and the shares may decline by 50% or more.
If, however, a low-growth company remains low growth but earns
returns better than the cost of capital, it will perform better than the
average.
SHAREHOLDER INDIFFERENCE
Managers have a choice about how to spend their earnings. They can
either distribute earnings to shareholders, or retain earnings to invest in
new activities. However, more simple modeling (Box 4.2 overleaf)
shows that if the new investments only earn the cost of capital (i.e., what
shareholders could earn on the market through reinvesting distributed
earnings), then shareholders should be indifferent. They will expect to
have the same result whether the company invests and continues to
grow, or distributes and forgoes growth (assuming the investments bear
a similar risk profile to that of the market). Shareholders are only inter-
ested in supporting growth ambitions if the new businesses are likely to
outperform the average company on the stock market.
Simple modeling also demonstrates that the total return for share-
holders is the same whether the low-growth company uses its spare
cash to buy back shares or pay additional dividends (Box 4.3).
75
Box 4.2 Shareholder indifference: growth vs.dividends
Take two companies, A and B, both with post-tax earnings of $500 million. Suppose they have
different strategies on dividend distribution. Company A distributes 90% of post-tax earnings
and Company B distributes 10%. Though Company B will grow faster than Company A, the
return to shareholders will be the same if the return from the new assets matches the cost of
capital.
Company A Company B
Post-tax earnings ($m) 500 500
Dividend distribution (%) 90% 10%
Dividends ($m) 450 50
Retained earnings ($m) 50 450
Post-tax return on new investments (%) 12% 12%
Additional post-tax earnings in first year ($m) 6 54
New level of earnings ($m) 506 554
New level of dividends ($m) 455.4 55.4
Dividend growth rate (%) 1.2% 10.8%
Cost of capital (%) 12% 12%
Market capitalization ($M) on basis of dividend

growth model, i.e.:
Market cap = Dividends/(Cost of capital – 450 / (0.12 – 0.012) 50 / (0.12 – 0.108)
dividend growth rate) = 450 / 0.108 = 50 / 0.012
So market capitalization ($m) is the same = 4166 = 4166
As is TSR%
= Growth in capitalization + dividends (4166*0.012+450) (4166*0.108+50)
Market capitalization 4166 4166
= 12% = 12%
These areas of shareholder indifference are not just theory. In July

2004, Microsoft announced the biggest ever program of returning cash
to shareholders. Dividends were increased to nearly 40% of current
earnings, a special dividend of $3 amounting to 10% of the market cap-
italization was paid, and share repurchases of up to $30 billion were
76
THE GROWTH GAMBLE
Box 4.3 Shareholder indifference: dividends vs.buybacks
Take two companies, A and B.Suppose both have a policy of full distribution of earnings, and so
have no growth from retained earnings. However, Company A pays out 100% of post-tax earn-
ings as dividends, and Company B uses earnings to buy back its shares. In both cases it is
assumed that earnings are retained through the year,and the share price rises accordingly in the
expectation that earnings will be distributed at the end of the year.
Company A Company B
Market capitalization at start of year ($m) 150 150
Expected post-tax earnings in first year ($m) 15 15
P/E ratio 10 10
Number of shares at start of year (m) 100 100
Share price at start of year (cents) 150 150
Share price at end of year just prior to distribution (cents) 165 165
Dividends per share (cents) 15 0

Number of shares repurchased (m) 0 9.1
Number of shares at end of year (m) 100 89.9
Expected earnings in second year ($m) 15 15
Market capitalization at start of year two ($m) 150 150
Share price at start of year two immediately post 150 165
distribution/repurchase (cents)
TSR%
= (Increase in share price + Dividends) 150 – 150 + 15 165 – 150 + 0
Share price at start of year 150 150
= 10% = 10%
declared over the next four years. In total Microsoft announced that it
would return $75 billion to shareholders. Yet the share price moved
very little. It had more than halved from $55 in late 1999 to $26 in
mid-2004, and it continued in the $25–30 range.
Managers might have presumed that a cash handout of this size
would be interpreted as an admission that Microsoft had few growth
prospects outside its core business. No doubt they agonized about the
77
risk of a further share slump. Evidently, the shareholders had figured

out that Microsoft was not going to be able to use this cash in its
growth plans (the $25 price presumed growth at about 10%) and had
factored this into their calculations. They appeared to be indifferent as
to whether the cash was returned to them or Microsoft was allowed to
manage it.
LIVING WITH MARKET EXPECTATIONS
HIGHLY RATED COMPANIES
The main problem that managers need to consider when expected growth
rates change—when a company hits a growth stall—is the shift in value
that occurs. A company with a growth track record is likely to be highly
rated, maybe with a price/earnings (P/E) ratio well in excess of 20. This
can present a particularly difficult problem. The current value is likely to
be based on an assumption of continued growth, at least in the medium
term. A decision by management to abandon attempts to get into new
businesses could cause the market to downgrade its assumptions about
growth, leading to a significant fall in the stock price. Rather than taking
the pain of a stock price fall today, managers are tempted to launch addi-
tional new businesses that at least have a chance of maintaining the growth
path and hence, if the shareholders believe the story they are told, main-
taining the share price. We call this problem the “overvaluation trap.”
In the short term, while market expectations remain fixed and a
company’s P/E ratio stays more or less constant, a company will
increase shareholder wealth so long as the post-tax return on retained
earnings exceeds the reciprocal of the P/E ratio. When P/E ratios are
high, companies can invest in relatively low-return investments and
still increase shareholder value. Take a simple example, of Company A
with a P/E ratio of 30 investing $100 million of retained earnings in a
new business. To justify spending $100 million of shareholder value,
the company need only earn 3.3% post-tax on a earnings multiple of
30 to match the original $100 million invested. Any new business earn-
78
THE GROWTH GAMBLE
Figure 4.2 Company A: Overvaluation trap
ing over 3.3% will add to shareholder value on a multiple of 30. This
gives management a considerable incentive in the short term to invest
in new projects to maintain earnings growth.
The position is illustrated by Figure 4.2.
If the new businesses succeed to such an extent that high earnings
expectations are met, then the share price will have been justified and
the company will continue to prosper—Outcome I. However, if the
new businesses only meet the cost of capital, then in due course the
market will rerate the company downward—Outcome III. In terms of
cumulative shareholder value, Outcome III will be the same as if the
company did not invest in the new businesses, but instead returned funds
to shareholders. The difference between investing in new businesses that
match the cost of capital and deciding to return cash to shareholders lies
in the timing of the share price adjustment. If cash is returned to share-
holders, the adjustment is likely to happen immediately. If cash is
invested in new businesses, the adjustment is likely to be delayed while
shareholders decide how well the new businesses will perform.
If the new businesses perform well and produce returns that exceed
the cost of capital, there may still be an adjustment downward. In fact,
there will be an adjustment downward if the new businesses earn a
79
return greater than the cost of capital but less than that required to
support the original P/E of 30. This is Outcome II—shareholders will
see a decline in share price, but they will be better off as a result of the
new businesses when compared with having the cash returned to them.
If, however, the new businesses fail to earn the cost of capital, the
fall from grace will be more precipitous—Outcome IV. The long-term
result will be worse than it would have been under the less risky policy
of returning the cash to shareholders.
For managers who know that growth in their core businesses is
likely to slow down sometime soon—for the managers in charge of
Intel and McDonald’s—the overvaluation trap presents a real dilemma.
Should they signal to the market today that growth will slow in the
future and face a significant, and possibly career-threatening, decline in
the share price? Or should they invest in new businesses, in a bid to
find new sources of growth, and risk a much more calamitous share
price decline if the new businesses fail?
The rational answer to the trap is to pursue new business projects
only if they have a reasonable chance of earning above the cost of cap-
ital. If insufficient new businesses pass the test, managers should signal
the change in growth expectations and aim for Outcomes II or III.
Given the failure rate of companies seeking additional growth from
new businesses, it is irrational to risk a calamitous share price decline
(Outcome IV) by making investments with a low probability of success.
However, managers are not wholly rational: they frequently fall into
the overvaluation trap. Both McDonald’s and Intel chose the high road
of aggressive investment in new businesses. Both efforts failed and con-
tributed to calamitous falls in the share price. McDonald’s appeared to
learn from the experience, announced lower growth ambitions and a
determination to focus on the core business, and was rewarded with a
significant improvement in the share price. Intel managers also
announced a more cautious new businesses policy—PC plus. However,
by 2004 the strategy had again become ambitious, with plans to enter
four or five new silicon businesses such as chips for flat panels and
mobile phones.
80
THE GROWTH GAMBLE
Figure 4.3 Company B: Undervaluation trap
LOWLY RATED COMPANIES
Companies that are lowly rated because of low growth expectations

face other issues. In many ways, theirs is the more enviable position, it
being easier to exceed low rather than high expectations.
So long as the P/E ratio remains the same, new investments will
only increase shareholder value if they earn more than the reciprocal
of the P/E ratio. Suppose Company B has a P/E ratio of 5 and invests
$100 million of retained earnings in a new business. To gain $100 mil-
lion of market capitalization the new business needs to earn $20 mil-
lion, assuming the P/E remains at 5 (5 times 20 is 100). Any new
business earning over 20% will add to market capitalization. A new
business earning less than 20% will reduce market capitalization. This
gives management a considerable hurdle in the short term. Thus, when
the P/E ratio is low, companies need to earn returns well above the cost
of capital in order to increase capitalization.
Figure 4.3 illustrates this position. Managers in Company B will be
tempted to try to rectify the undervaluation through greater distribu-
tion of earnings via dividends or share buybacks. This ought to result
81
in a rerating of the company as an income stock. If the average return

shareholders can earn elsewhere is 12%, then, as an income stock,
Company B should be rerated upward to a P/E ratio of 8.3—Outcome
II.
Investing in new businesses may delay the rerating because the new
investments will not prove their potential for a year or two. The delay
will only be worthwhile if the return on the new investments is in
excess of the cost of capital—Outcome I. The “undervaluation trap,”
therefore, is that managers can feel trapped into low-growth, high-
distribution strategies, even though there may be good opportunities
to invest in new businesses.
If the new initiatives only meet the cost of capital, then the final
outcome will be the same as distributing the reinvested funds to share-
holders—Outcome II—but with a delay in its achievement. However,
if the new businesses fail to meet the cost of capital, then Company B
can expect to remain lowly rated—Outcome III.
The rational answer to the trap is to invest only in new businesses
that have a reasonable chance of earning above the cost of capital.
However, managers often choose the low-risk path: they announce a
back-to-basics strategy with a high dividend or share buyback as a way
of restoring confidence. Then, once the company is fairly valued, they
consider investments in new businesses. Choosing not to invest, and
opting for greater distribution, delivers a very attractive immediate
total shareholder return; in the case of Company B, a TSR of 20%.
With earnings of $100 million and a P/E ratio of 5, the company will
have a market capitalization of $500 million. Buying back $100 mil-
lion worth of shares, or distributing the full $100 million in dividends,
gives a 20% return. Whereas it is easy to criticize managers for mis-
handling the overvaluation trap by overinvesting in risky projects, it is
harder to criticize them for choosing the risk-averse strategy when
faced with the undervaluation trap. However, if the company does
have good new business opportunities, the value lost can still be
significant.
82
THE GROWTH GAMBLE
WHEN IS LOW GROWTH THE RIGHT CHOICE?
Low growth is the right choice for companies under two conditions:
❏ When there are few opportunities for the core businesses to grow.
❏ When there are few new businesses that have a good chance of out-
performing the average company on the stock market.
Companies following a low-growth strategy under such circumstances

will minimize their risks of value destruction and increase their chances
of value creation.
When deciding on a company’s growth policy, the following princi-
ples should guide the choice:
❏ Companies can only add to shareholder value if their growth

exceeds market expectations—just growing is not enough. High
growth is an expectation treadmill.
❏ When new business opportunities earn the cost of capital (roughly
the average return on the stock market) shareholders are indifferent
between making the investment or returning the cash through div-
idends and buybacks. Thus investments in new businesses need to
have a good chance of exceeding average returns to justify manage-
ment attention.
❏ If a company cannot find new business projects that are likely to
earn more than the cost of capital, surplus funds should be returned
to shareholders.
❏ Low-growth companies can still produce good returns to share-
holders.
❏ So long as there are no tax differences between income and capital
gains, shareholders are indifferent between dividends and share
buybacks.
We have also looked at the particular dilemmas faced by companies

with high and low P/Es. Companies with high P/Es face an over-
valuation trap:
83
❏ In the short term, highly rated companies may do better for their
shareholders by making new investments that make quite low
returns. However, unless the returns are sufficiently high to warrant
the original high rating, the share price will eventually adjust down-
ward to reflect the returns actually received.
❏ So long as the returns from new businesses exceed the cost of capi-
tal, shareholders will be better off if the company invests: even
though the share price may fall, it will fall less than it would with-
out the investments.
❏ The fall in share price will be even greater if the new businesses fail
to earn the cost of capital.
❏ Thus the rational stance for companies that cannot see ways of
maintaining high expectations of growth is to pursue new business
projects only when they are likely to earn more than the cost of cap-
ital and manage a decline in their share price to reflect the appro-
priate value.
Companies with low P/Es face an undervaluation trap:
❏ In the short term, new business investments will need to earn con-
siderably more than the cost of capital to add to immediate share-
holder value.
❏ Increasing distributions, whether by dividends or share buybacks, is
a tempting, low-risk option that can give very good immediate
shareholder returns, and could lead to a rerating upward as a good
income stock.
❏ However, shareholder value should be increased if the new busi-
nesses outperform the average, and exceed the cost of capital.
❏ Nevertheless, the low rating will remain if the new businesses fail.
❏ Thus the rational stance for undervalued companies that see good
opportunities is to pick only those projects that have a high chance
of success, recognizing that paying back funds to shareholders will
be a more attractive alternative than making average investments.
84
THE GROWTH GAMBLE
CAN LOW GROWTH BE A WINNING STRATEGY?
Despite the logic of this chapter, most managers remain unconvinced.

Growth has been seen as a necessary objective for so long that man-
agers find it hard to accept low growth. They feel certain that low
growth will lead to problems that simple financial theory has over-
looked. They feel insecure without a growth plan.
To support our view that low growth is not only sensible for many
companies but also financially attractive, we would like to offer some
examples.
CROWN CORK AND SEAL
Probably the most famous example of a company succeeding by

taking a lower-growth, more focused strategy is Crown Cork and
Seal.
In 1891, a manager in a Baltimore machine shop hit on an idea for
a better bottle cap—a piece of tin-coated steel with a flanged edge and
an insert of natural cork. The new product was a hit and the company
prospered until patents ran out in the 1920s. In 1927, Crown Cork was
bought by a competitor, Charles McManus.
McManus steered the company to a dominant position in the world
market for bottle caps. He also anticipated the rise of the beer can and
built a one million square foot plant in Philadelphia. McManus’s suc-
cess, however, did not survive his death in 1946. The Philadelphia
plant proved to be a failure because it was not close to its customers,
and his successor invested unsuccessfully in other new businesses such
as plastics and bird cages.
By the mid-1950s the company was in distress, and John Connelly,
a local businessman, shareholder, and previous unsatisfied customer,
became president. Connelly’s plan was “just common sense.” He
removed layers of unnecessary management, cut staff functions in half,
closed central research, shut down the Philadelphia facility, invested in
plants close to customers, and gave each plant manager responsibility
for profitability, quality, and customer service.
85
Faced with much larger competitors—American Can and

Continental Can—Connelly decided to focus on crowns, where he was
still market leader, and specialist segments of the canning industry. He
developed a strong position in motor oil cans and in cans for “hard-to-
hold” liquids, such as soft drinks and aerosols.
As the market for cans matured in the 1970s and 1980s and large com-
panies, like Campbell’s and Coca-Cola, began to make their own cans,
Crown’s competitors invested in new packaging solutions with new mate-
rials such as fiber-foil, paper, and plastic. Their vision was to convert
from being canning companies to becoming packaging solutions compa-
nies. Connelly, however, chose to go firmly in the opposite direction.
Crown focused on crowns and cans, expanding internationally using
secondhand equipment, helping customers convert from steel to alu-
minum, and investing in plants close to customers’ canning lines. Since
the business was highly profitable, Connelly was quickly able to pay off
the debts he inherited. He then paid off all the preferred stock and,
from 1970 onward, he used spare cash to buy back shares.
In the 1980s margins came under pressure. However, Crown’s high
levels of depreciation generated sufficient cash to support a modest
level of real growth and buy back significant numbers of shares. In
1970 there were over 20 million shares. By 1987, Connelly had repur-
chased more than half. Not surprisingly, earnings per share performed
especially well, rising from $1.41 to $9.28 in 1987. Similarly, the share
price rose from less than $5 to nearly $50.
In the meantime, Crown’s competitors had concluded that cans
were a mature market. Not only did they expand into other growth
areas of packaging, but they also invested in other growth sectors,
including insurance and oil. Unfortunately, both kinds of diversifica-
tion produced low returns. There was excess capacity in the growth
areas of the packaging industry because most packaging companies
were entering the new markets. In addition, the non-packaging busi-
nesses performed poorly, due in part to the lack of experience of man-
agers schooled in the canning industry.
By the mid-1980s, American Can, National Can, and Continental
Can had all been the target of aggressive acquirers looking for under-
86
THE GROWTH GAMBLE
performing businesses to turn around. In 1985 American Can was

acquired by Triangle Industries, who also acquired National Can the
following year. In 1988 the integrated company, American National,
was sold to Pechiney, a French aluminum company.
The Continental Group was acquired in 1984 by Peter Kiewit, who
sold most of the non-packaging businesses as part of a program to
restore focus and profitability. When Kiewit was looking for a buyer of
the improved business, an obvious candidate was the best-performing
company in the can industry: Crown Cork and Seal.
Having avoided diversification outside metal crowns and cans for
more than 20 years, Crown, with one huge deal, not only became one
of the largest metal packaging companies in the world, but also added
plastic, glass, and foil packaging skills. Its timing was both good and
bad. The deal was completed near the top of the 1980s boom, causing
Crown to pay a full price. However, the company’s entry into other
areas of packaging was timed to suit its skills. By the early 1990s the
“new packaging materials” were beginning to mature into predictable
businesses that would respond well to Crown’s cost-conscious, decen-
tralized, customer-centric management style.
If the tale could be ended at this point, Crown Cork and Seal would
be the ideal example: a company that ended up dominating its industry
while its competitors became distracted with new businesses. However,
the story did not stop there. The growth pressures of the 1990s
seduced even this focused management team. Crown continued to
make acquisitions, adding further businesses in Europe and other parts
of the world to become a leader in metal containers and one of the
world’s top packaging companies. In the process it paid too much for
its acquisitions and the business began to get out of control. Toward
the end of the 1990s the company’s profits collapsed and its market
capitalization fell by 90%. In the end, the pressure to grow and the
ambition to become the world’s leading packaging company undid
much of the work of John Connelly’s low-growth strategy.
87
IBM
Metal packaging is a mature industry and Crown’s experience demon-

strates the value of focus in an environment that is not growing. IBM, on
the other hand, has been in a fast-growth, high-tech industry. Yet like
Crown, IBM has demonstrated the benefits of limiting growth ambitions.
Under Louis Gerstner IBM delivered good returns to shareholders
through modest growth combined with aggressive share buybacks.
Between 1993 and 2003, it delivered a cumulative total shareholder
return of 22% per annum while distributing nearly all its earnings back
to shareholders.
Gerstner joined IBM on 1 April 1993 at a time of extreme crisis,
when its share price dropped below $50 compared to over $175 in
1986. That summer it fell further to an all-time low of $41. Dividends
had been cut drastically in 1992, from $4.84 to $2.16, and then by
Gerstner himself to $1 in July 1993. Yet by March 2004 the share price
had risen to an equivalent of $367 (taking account of two intervening
2:1 scrip issues). Over that period total revenue grew by an equivalent
of only 1.9% per annum in real terms, well below US GDP, which
grew by an average 3.2% over the same period.
How did Gerstner do it? First, he offset the decline in the hardware
business by building a major services business. But this did not give
him significant growth. He therefore needed a second strategy to get
earnings per share and the share price rising. This was his share buy-
back program.
IBM had bought back shares previously. Between 1986 and 1989 it
bought back around 6% of issued shares, spending over $4 billion.
However, Gerstner made share buybacks a central strategy. Between 1995
and 1998 he bought 29% of outstanding common stock, spending a net
$22.2 billion. This was more than 100% of post-tax earnings over that
period of $21.9 billion. Dividends were also increased, but only from $1.00
to $1.88 per share. In all, from 1995 to 1998, IBM distributed on average
114% of earnings to shareholders by way of dividends and share buybacks.
By decreasing the number of shares, a 33% increase in earnings
from 1995 to 1998 turned into an 87% increase in earnings per share.
88
THE GROWTH GAMBLE
The share price increased by 152%, being boosted by a rising P/E ratio
over the period, from 13.2 to 17.8. The pace of share purchase and
retirement eased slightly after 1998. Even so, over the 10 years 1994 to
2003, revenue increased by an equivalent of 3.6% per annum, earnings
by 11.1% per annum, earnings per share by 15.1% per annum, and the
share price by 21.4% per annum.
During this period IBM did not stop developing its businesses, mak-
ing two significant acquisitions: Lotus Corporation for $3.2 billion in
1995, and the consulting arm of PricewaterhouseCoopers in 2002 for
$3.5 billion. It also continued to invest in its core businesses, not least
through its significant spend on R&D and engineering. This averaged a
pretty steady $5.1 billion per annum. Thus over the period 1993–2003,
IBM spent $56.2 billion on R&D, earned net of tax $51.4 billion, and
spent $52.6 billion (gross) buying back its own shares ($44.3 billion net).
Would shareholders have been better off if IBM had invested more
of its earnings in new businesses rather than its own shares? Many
commentators advised this over the period, not least during the heady
dot-com days. IBM, after all, was close to the center of a technological
revolution. Gerstner’s view was different. He credited much of his suc-
cess to the businesses he did not enter. He told a Business Week
journalist:
If life was so easy that you could just go buy success, there would be a
lot more successful companies in the world. There were few PC com-
panies that we weren’t offered at some point. Telecommunications
companies all over the world proposed joint ventures. There were peo-
ple who suggested we get into the content business. We turned them
all down.5
Instead of investing in these new opportunities, Gerstner chose a

lower-growth path. As we have outlined, he continued to invest heav-
ily in IBM’s core businesses through R&D, made two significant, but
not huge, acquisitions that added to existing growth businesses, and dis-
tributed over 100% of earnings back to shareholders. The result was a
performance well ahead of most of his growth-driven rivals. He argued:
89
The value that some analysts put on revenue growth versus what they
put on profit is out of whack. There is a huge difference between what
we hear from our owners and from analysts and the media. Every time
I met with owners they said “Please don’t take the cash and invest it in
stupid acquisitions or ventures.”
With the handover to Palmisano, IBM has become more growth ori-
ented. The Emerging Business Opportunities program described in
Chapter 7 is one process that Palmisano has put in place to try to accel-
erate growth. We believe that this has helped remove some of the cul-
tural and structural barriers to new businesses inside IBM and may
make some small addition to the company’s growth rate. But we are
concerned that IBM’s increasing focus on growth will cause managers
to be less selective about which businesses to invest in. Because
Palmisano’s growth goals appear to be driven more by ambition than
by a realistic assessment of prospects for new businesses,6 we are con-
cerned that IBM may make the same mistake as Crown: drive for
growth without sufficient caution.
DO LOW-GROWTH COMPANIES EVER BECOME

HIGH-GROWTH COMPANIES?
Part of the reason managers are so uncomfortable setting low growth

targets is that they fear the long-term implications. They fear that once
their company settles for low growth it will never be able to change
gear and become a high-growth company. So far as Fortune 50 com-
panies are concerned, this appears to be true. As we have noted, only
4% succeed in restarting growth after a stall.
Armed with this knowledge, the Corporate Strategy Board decided
to see if the same were true in smaller companies.7 They looked for
companies whose “revenue growth performance began a sustained and
meaningful positive inflection from its past performance,” a phenome-
non they called “growth restart.” From a sample of 1,487 large mature
firms, they identified 490 whose three-year revenue growth most out-
90
THE GROWTH GAMBLE
performed competitors; 174 of these companies also had growing mar-

gins. Of these 174, 30 were labeled “continuous growers” and 144
“growth restarters.” In other words, some 10% of companies do suc-
cessfully restart growth following a period of slower growth. This is a
more promising number than the 4% in the Fortune 50, but it is still a
small percentage. While it may depress many managers of mature
companies, it underlines the importance of being cautious about
investing in new growth ventures when the core businesses are mature.
Since most companies try and less than 10% succeed, the net accumu-
lated effort is likely to be value destroying.
Further analysis by the Corporate Strategy Board should cause
managers to be even more cautious. Of the 144 growth restarts only 42
were “sustained restarts.” These were companies that managed to sus-
tain their growth rate at levels higher than the industry average. In
other words, less than 3% of mature companies were able to restart
growth and sustain rates above the average of their industries.
Unfortunately, the statistics get even worse: 60% of these sustained
restarts achieved their success by revitalizing their existing businesses.
Only 40% succeeded by developing new businesses, and less than half
of these, only 11 companies, succeeded by developing new businesses
outside their core. In other words, less than 2% of companies restart
growth through developing new businesses and less than 1% do it by
finding opportunities outside their core.
While these statistics reinforce management’s desire to avoid letting
their companies reach maturity, they also underline the main message
of this book: that there are few opportunities for mature companies to
create significant new businesses. Managers who attempt to develop
new businesses need to recognize that their success does not depend on
the amount of effort they put in. It depends on whether the combina-
tion of market opportunity, competitor behavior, internal skills and
resources, and leadership that is needed for success exists at a particu-
lar point in time. For many companies at many points this combination
does not exist, and these companies would be better keeping their pow-
der dry. For some at some points, the combination does exist, and these
companies should invest with confidence.
91
CHAPTER 5
THE NEW BUSINESSES

TRAFFIC LIGHTS
T
he central message of this book is that managers need to be
more selective about the investments they make in new growth
businesses. In our view the pressures to grow, the natural opti-
mism of managers, the over-reliance on the financial business case, along
with a number of the reasons given in Chapter 3 cause managers to be
insufficiently thoughtful and rigorous when selecting new businesses in
which to invest. The numerous failures of both new business initiatives
and acquisitions by many of the major corporations around the world
attest to this. Much shareholder value has been needlessly destroyed.
This chapter, which describes our selection screen, is therefore the most
important in the book. If managers understand the logic behind the
Traffic Lights and use its criteria to choose projects, they will invest only
in those projects that have a reasonable probability of success.
This chapter alone is not enough. Managers also need to have read
Chapter 4 on low growth. Without understanding the financial conse-
quences of low growth or acknowledging that many companies have
long periods of low growth, managers will not be comfortable using
the Traffic Lights to screen out most of their projects. Believing that
low growth is unacceptable, they will either be deliberately optimistic
about some of the Traffic Lights or decide to invest in a new business
even when the Traffic Lights suggest that the chances of success are
too low. Unless managers understand that low growth is acceptable,
they will be tempted to gamble.
The Traffic Lights have been developed out of our research. We
examined the criteria managers use to evaluate new businesses. We
THE GROWTH GAMBLE
Figure 5.1 The New Businesses Traffic Lights
shadowed managers in the process of screening new business ideas. We

studied the criteria suggested by other authors. Finally, we applied
pure logic. The tool (Figure 5.1) has taken nearly three years to
develop and test.
The Traffic Lights involve making red, yellow, or green judgments
about four questions:
❏ Do we have a significant advantage (green), small or uncertain advan-

tage (yellow), or significant disadvantage (red) in this new business?
❏ Is the profit pool for this new market average (yellow), a “rare
game” (green), or a “dog” (red)?
❏ Do we have leaders of this new business (and sponsors in the parent
company) that are clearly superior to (green), similar to (yellow), or
less strong than (red) competitor businesses?
❏ Is the impact of this new business on existing businesses likely to be
significantly positive (green), uncertain (yellow), or significantly
negative (red)?
The purpose is to decide whether the project should go ahead or not.

We have found that any one green signal can be enough to make the
project into a go, so long as there are no red signals. One red signal is
93
THE NEW BUSINESSES TRAFFIC LIGHTS
enough to stop the project. A project with all yellow signals is

marginal.
When HP considered entering the personal computer business in
1981, for example, three traffic lights were green:
❏ HP had reason to believe that it could gain a technical advantage in

large computers for commercial applications because mainframes
were threatened by the new client-server model, IBM was closing
down its RISC development effort, and HP could acquire IBM’s RISC
team. By moving to RISC technology, HP could leapfrog incumbents
focused on the CISC software used by their installed base.
❏ The profit pool looked exciting: most manufacturers were making
attractive returns.
❏ HP also had a strong team running its computer business and suit-
able sponsors at the corporate level. It had been manufacturing
minicomputers for many years and selling them mainly into techni-
cal applications. By chance, the management team in charge of the
business was particularly strong. On several previous occasions they
had proposed entering the market for general-purpose commercial
applications and had been turned down.
On the other hand, when British Sugar, a company focused on pro-

cessing sugar beet, looked at the opportunity of entering the hemp
processing business, four traffic lights were red:
❏ An advantage would be hard for British Sugar to create. The com-

pany British Sugar was thinking of acquiring was not the best player
in the market and British Sugar had limited skills to bring to this
new area.
❏ The market appeared to be a “dog.” The auto industry and matting
customers who bought most of the hemp had a history of providing
low margins. Even if changes in legislation caused significant
growth (the main reason for considering this opportunity), the
prospect of earning good returns was limited because the main
customer would be the auto industry, the toughest of all buyers.
94
THE GROWTH GAMBLE
❏ The team British Sugar could assemble from the acquisition and
internal appointments would be less experienced and less customer
led than competitors.
❏ Finally, the impact on the existing business would be likely to be
negative. The project would involve some of the most able man-
agers from both the processing and technical sides of British Sugar
at a time when these managers could be much more profitably
employed in the sugar business, driving costs down and developing
new service offerings for customers.
The New Businesses Traffic Lights cover many of the criteria that are
commonly used to screen projects. For example, Tim Hammond, then
corporate development and group marketing director at Whitbread,
used a simple matrix (Figure 5.2) to screen 30 ideas for new leisure
businesses. Whitbread is a restaurant, hotels, and leisure group and was
exploring opportunities in new restaurant areas, movie cinemas, spas,
wellbeing services, sports clubs, and other leisure activities. In fact it
was from Tim Hammond that we picked up the idea of a traffic lights
system. He defined criteria for each axis of the matrix and graded each
Figure 5.2 Potential/fit matrix
95
criterion red, yellow, or green. By eliminating all those with a red sig-
nal, he reduced the number of options to six. Further investigation
identified only two with multiple green signals.
The difference between the Traffic Lights and the criteria used by
most managers, such as Tim Hammond, is in the detail. The Traffic
Lights are built on some fundamental insights about the circumstances
that lead to value creation and the mistakes that managers make when
evaluating new businesses. In total there are five insights, ranging from
the realization that managers often underestimate learning costs to the
observation that they often overlook the costs of distracting attention
from existing businesses. Box 5.1 summarizes these five insights. Their
impact is to highlight some negative criteria that are often ignored or
given insufficient weight. The result is that the Traffic Lights reject
more projects than most normal screening processes.
The rest of this chapter examines each of the Traffic Lights in more
detail. To illustrate how the tool can best be used we will focus on two
situations: whether a chocolate confectionery company such as Mars or
Cadbury should enter the ice-cream business in Europe using its
brands and its experience with snack products; and whether a retailer
of health and beauty products, such as Boots in the UK or Neiman
Marcus in the US, should set up a chain of eye-care shops using its
branding and retailing skills.
This is a long chapter. It is, therefore, worth providing some guidance
to readers. Those who want to focus on the unique ideas in this chapter
need only read the five insights (in bold) and the text that relates to
each—about 11 pages in total. Those who have a new business idea that
they want to assess will need to read enough under each heading to be
confident that they understand the criteria being used and how to make
the assessment—probably more than half of the pages that follow. Finally,
those who already have some criteria for selecting new businesses and
want to see whether the Traffic Lights contain additional ideas should
focus on those sections that are unfamiliar to them—probably a third or
a quarter of the chapter. In other words, it is unlikely that any reader will
want to work through all of the next 50 pages in sequence. Box 5.2 over-
leaf provides an overview of all the criteria used in the Traffic Lights.1
96
THE GROWTH GAMBLE
Box 5.1 Five insights that inform the Traffic Lights

Insights Reasons Implications
1. Managers do not normally The value reason for entering a Since many types of unique
consider the tradability of their new business is the extra value value, such as brand or patent,
unique value when assessing that can be created. Hence man- can be licensed or “cashed”
new businesses (page 102). agers should deduct from their through a joint venture,the value
calculations that part of their advantage equation is often neu-
unique contribution that can be tral (yellow) or negative (red).
turned into value without enter-
intg the new business.
2. Managers do not normally Because learning costs are hard Since learning costs are rarely
assess the likely costs,at both the to quantify they are usually left less than 10% of profits and can
operating level and the corporate out of the equation. Fit analysis, be 50% or more at least for a few
level, of learning a new business one way of identifying likely years, the value advantage equa-
(page 110). learning costs, is normally tion is often neutral (yellow) or
focused on what fits rather than negative (red).
on what does not fit.
3. Analysis of markets should Strategy analysis demonstrates Since most markets are at neither
focus on identifying extreme situ- that in most markets companies extreme,they score yellow on the
ations that are either so good that with an advantage will earn Traffic Lights, offering no infor-
even a competitor with disadvan- above-average returns. Hence mation for or against the strate-
tage can earn a good return or so detailed analysis of growth rates gic decision.
bad that even an advantaged and likely margins, while neces-
player may earn less than the cost sary for a financial plan, is often
of capital (page 117). unnecessary for making a strate-
gic decision.
4.Managers do not normally give Managers, especially in large When an objective assessment is
sufficient attention to the issue companies, believe that their made of the quality of the pro-
of who is going to run a new managerial resources are consid- ject’s managers and sponsors
business and, particularly, who erable or that good talent can be compared to those of current and
the new business is going to hired from the marketplace.They likely competitors, it is often evi-
report to (page 130). also presume that they can learn dent that they are inferior (red)
most new businesses. or at least not superior (yellow).
5. Managers often underestimate Once managers have decided Most new businesses create some
the loss of performance in the that the core is not enough they distraction from existing busi-
core businesses that occurs when often switch too much of their nesses. Distraction costs become
attention shifts to new businesses attention to the new. Moreover, significantly negative (red) when
and some of the most energetic new growth businesses often the new business competes with
managers are allocated to new offer more attractive careers for existing businesses for scarce
business projects (page 142). the company’s better managers. resources or skills.
97
Box 5.2 The New Businesses Traffic Lights
Value advantage
❏ Our unique contribution (at operating and parenting levels)
❏ Less % of our contribution that is tradable
❏ Less unique contribution of competitors
❏ Less cost of learning the new business (at operating and parenting levels)
Profit pool potential

❏ Business model potential for high margins (value relative to cost, break even as % of mar-
ket)
❏ Industry structure potential for high margins (five forces taking account of likely growth
rates)
❏ Opportunity for us to be a leader in this market
❏ Cost of trying relative to size of profit pool (taking account of time to commercialization)
❏ Business model vulnerability (number of enablers, sensitivity to key variables)
Leadership/sponsorship quality
❏ Relative quality of MD/leadership team of the unit
❏ Significance of MD/leadership’s personal insights about the business
❏ Status of sponsor within main parent
Impact on existing businesses

❏ Size of positive or negative synergies
❏ Risk of distraction costs
DO WE HAVE A VALUE ADVANTAGE?
One of the truths of business is that a superior return on capital is

normally the result of some area of advantage. Success is hard to
achieve without an advantage. Michael Porter captured this thought
compellingly in his book Competitive Advantage.2 Business strategy, he
argued, should be about how companies can achieve advantage over
their rivals. Advantage, he proposed, comes either from superior costs
or from superior products.
98
THE GROWTH GAMBLE
Box 5.3 Value Advantage Equation
Unique value we bring (from the operating and parent level)

Less Percent of value we could trade (through sale, license, or joint venture)
Less Unique value our competitors bring
Less Cost of learning the new business, at the operating level and at the parent
level, relative to competitors
Equals Our value advantage
(green if significantly positive, red if significantly negative)
Many managers, consultants, and academics have built on Porter’s

ideas and defined hundreds of different ways of creating advantage,
including the design of the business model, market share, special com-
petencies, close relationships, market positions, and unique assets.
In particular, the resource-based view of strategy, developed about
the time of the publication of Porter’s book, focuses on the resources
that companies possess as the main explanation for performance dif-
ferences. The most important resources, the theorists argue, are those
that are valued by customers, rare and difficult for rivals to acquire or
imitate. Competencies that are “path dependent,” meaning that man-
agers need to have had a certain set of experiences to have developed
the skill or knowledge, are often especially valuable because they are
hard for others to acquire or copy.
So how does a management team assess whether it is likely to have
an advantage in a new business situation? Clearly, this is difficult. It can
be especially difficult if the advantage is based on competencies that are
hard to understand. As a result, we have developed the Value Advantage
Equation to help managers make the right judgment (see Box 5.3).
WHAT UNIQUE VALUE DO WE BRING?
The equation starts with an assessment of what the company can con-
tribute to this new business that has special or unique value. Figure 5.3
overleaf lists some of the areas of special contribution that we
99
Figure 5.3 Sources of unique value
identified from our database of success stories. Most of the successes

were based on more than one source of unique value, the average being
two. Just under 20% of the successes were also influenced by luck. In
these cases the sources of unique value were fewer, averaging only 1.5.
We considered trying to develop a checklist that managers could
use to define their unique contribution, but decided against it. The
checklist could never be complete and it might encourage managers to
tick boxes rather than think through their unique situation. Managers,
we decided, need to define as precisely as possible what special contri-
butions they believe they can make to the new business and how valu-
able the contributions are. It is the judgment about value that is the
hardest.
For example, the health and beauty retailer considering eye care
might have a list of special contributions like this:
❏ Trusted health and beauty brand.

❏ Excess retail space in some locations.
❏ Knowledge of the health and beauty consumer.
❏ Distribution system to every significant retail location.
❏ Retailing skills.
❏ Employee loyalty.
100
THE GROWTH GAMBLE
The key question is “How unique?” How much advantage does each
contribution deliver?
The brand could clearly be an important contribution. If the brand
could drive extra sales volume to each site compared to existing brands
in the market, it could significantly increase profitability.
The excess retail space might be a valuable contribution, if it were
especially valuable to the eye-care business. Unless we are confident
that the company has a number of sites that are especially good for eye-
care retailing, we would have to assume that the spare retail space
would be useful but not a significant contribution.
Knowledge of the health and beauty consumer could be important,
but again it is unlikely to be a unique contribution. Other competitors
have been in the market for many years. Unless the health and beauty
managers believe they possess some insight about this consumer seg-
ment that eye-care competitors do not have, the knowledge is unlikely
to be a significant contribution.
The distribution system could be a significant contribution if it
would enable the new eye-care business to have fewer lines out of stock
or have lower supply costs than competitors. Given the maturity of the
existing industry and the availability of third-party distribution serv-
ices, this is unlikely to be a source of significant advantage.
Retailing skills and employee loyalty could be a major contribution.
This would depend on whether the application of these skills would be
likely to enable the business to serve more customers, attract more cus-
tomer loyalty, operate at lower cost, or sell at higher prices than com-
petitors. The company might well believe that this would be a
significant contribution.
In summary, the health and beauty managers might have concluded
that they had an advantage in branding and in retailing skills, which
might enable them to earn significantly higher margins than most
competitors, assuming all other aspects are equal.
To make best use of the Value Advantage Equation, managers need
to convert judgments like “significantly higher margins” into a number
they feel is realistic, such as 50%. When faced with a number, vague
words like significant become less misleading. In ideal circumstances,
101
the analysis would involve reverse engineering the business models of

important competitors. These would then be compared to the business
model of the proposed new business to see how big the value advan-
tage is likely to be. With some detailed reverse engineering, it is
possible to make quite fine-grained judgments, distinguishing for
example between a 30% and a 40% advantage.
Assessing the unique contribution and putting some value on it,
such as “50% higher margins,” is easier when talking about an existing
market, such as eye care. When the business is a new market, for exam-
ple internet service provision in the early 1990s or outsourcing auto
safety testing in 2000, the task is harder. Managers need to compare
their special advantages with those of likely future competitors and try
to identify those unique resources that will be a significant source of
advantage in five years’ time. Guesswork comes into play along with
the biases of the person doing the guessing. Nevertheless, laying out
the value that managers believe is unique and assessing its significance
to relative profitability forces managers to make the necessary strategic
judgments.
WHAT VALUE CAN WE TRADE?
Insight 1: Managers do not normally consider the tradability of

their unique value when assessing new businesses.
The branding and retailing skills of the health and beauty company
may be valuable resources. One way to turn these resources into value
is by entering the eye-care business. However, there is another way for
the health and beauty company to turn these resources into value. It
can try to sell or license them, either to existing competitors in the eye-
care business or to some new entrant.
For example, if the brand increases sales by 10%, resulting in an
increase in margin from 10% to 15%, then it ought to be possible to
go to one of the less successful players and offer to license the brand.
While it would not make sense for the competitor to pay as much as a
5% royalty, it might be possible to earn a 2.5% or 3% royalty just by
102
THE GROWTH GAMBLE
licensing the brand. In this way, the health and beauty company could
get a proportion of the value of its resource without any of the risks and
costs associated with entering a new business. Disney, for example,
routinely licenses its brand and characters to other companies.
In the Value Advantage Equation, we deduct the tradable propor-
tion of the unique value from the original contribution to arrive at a
non-tradable contribution. The reason for this deduction needs care-
ful explanation.
We are trying to assess the logic for entering this new business. As
will become clear later in the Value Advantage Equation, there are costs
and risks associated with entering a new business. In order to justify the
costs and risks, we need to believe we have a contribution to make that
can only be turned into value by incurring these costs. If we can get
value for our contribution without incurring the costs, we have no rea-
son to enter the new business. We would be better advised to cash in
our contribution, saving ourselves the costs. In other words, it is the size
of the non-tradable portion of our unique contribution that is the key
reason for going ahead. The focus on non-tradable contribution is sup-
ported by our sample of successes. The vast majority had significant
sources of value that could not be traded (Figure 5.4 overleaf).
Another example might help. Transco, a gas pipeline business, had
underground channels and rights of way that were valuable to any
company in the business of laying fiberoptics for the internet and tele-
coms industry. The channels and rights of way reduced the costs and
increased the speed of laying communication fiber. This valuable
resource was causing Transco to think about entering the telecoms
business: Transco could lay fiber and sell the communication capacity
to its existing gas customers.
An alternative option for Transco would be to lease the channels
and rights of way to an existing telecoms company, either for a fixed fee
or for a share of the revenues. When deciding whether to enter the
business, Transco managers needed to compare entering the new busi-
ness with the alternative of leasing the resource to someone else. At the
time, they concluded that they could not get full value from leasing the
rights of way and chose to enter the business. With hindsight this may
103
Figure 5.4 Tradability of unique value
have been the wrong judgment, since Transco lost a significant amount
of money when the telecoms industry collapsed. Williams, a US gas
pipeline company, faced the same choice and made the same decision.
However, because Williams could lay fiberoptic cable at a cost 20%
below the industry average and because this source of advantage would
have been hard to trade for value, Williams probably made the right
decision.
In practice, of course, there are many alternatives, including different
kinds of leases and different kinds of joint ventures, involving different
parts of the business. For example, the pipeline company could lay the
fiber under an agreement with a telecoms company who would lease the
fiber. The price paid could be on a sliding scale dependent on the vol-
ume of traffic. An agreement could be made to lease back some capacity
for use by the pipeline company and for selling on to gas customers.
In an ideal world all these different options would be defined and
evaluated. However, options normally come on the table one at a time.
In practice, the managers in Transco put together a proposal for enter-
104
THE GROWTH GAMBLE
ing the telecoms business in joint venture with an experienced tele-

coms company. At this point the deal on offer needed to be assessed.
Rather than sending the managers back to lay out all the other possi-
ble options, we have found it best to deduct from the unique value the
amount that managers believe they could release through a licensing or
trading relationship. Discounting the value is a realistic way of taking
account of the alternative options.
Returning to the eye-care example, we should deduct something for
the tradability of the brand. If managers believe that it would be possi-
ble to license the health and beauty brand to one of the current com-
petitors, we should deduct the value of a likely license fee, say 2.5%.
We should also consider deducting some of the value attributed to
retailing skills and employee loyalty. However, experience suggests that
the company is unlikely to be able to trade its retailing skills and
employee loyalty for any significant value. Even if it were successful in
offering these skills as consultants, it is unlikely that it would create any
significant value contribution.
The non-tradable contribution, therefore, is the additional value
the company could get by using the brand itself plus the value it can
add through its retailing skills and employee loyalty. If the previous
estimate was a 50% higher margin, the non-tradable contribution
might be more like a 25% higher margin.
Managers often criticize our process at this point, arguing that it is
impossible to put figures on something as qualitative as non-tradable
contribution. Our response comes in two forms. First, we point out
that in acquisition situations these issues have to be quantified in order
to arrive at a price the buyer is prepared to pay. The unique value in an
acquisition is the “synergy.” If it is possible to estimate synergies in an
acquisition, it should be possible to estimate the equivalent in a new
business situation. Our second response is that accurate quantification
is not necessary. All that is necessary is to judge whether the non-
tradable contribution is larger than the combined value of the other
terms in the Value Advantage Equation. We recognize that the judg-
ments required are hard, but we believe that the format of the Value
Advantage Equation helps.
105
WHAT IS COMPETITORS’ UNIQUE VALUE?
The flip side of assessing our unique value is to consider what unique
value the competitors bring to the business. The analysis can focus first
on the existing competitors, but should focus also on potential
competitors.
The list of contributions of existing competitors in the eye-care
business might have been:
❏ Low rents in some locations due to leases signed some years before
and/or property purchased at lower prices.
❏ Established locations and knowledge about location advantage.
❏ Volume advantages in purchasing costs and other costs not con-
nected to the site.
❏ Brand strength, customer loyalty, and switching costs.
❏ Knowledge of customer preferences.
In the same way as before, these contributions need to be assessed in

terms of their worth: how much advantage these contributions give.
The low rents will give a calculable cost advantage. However,
should this cost advantage be taken into account? The competitors
ought to be assessing their property costs against market values (the
rent they could get if they sub-leased the property), hence their central
property company ought to be charging a market rent for the sites. If
so, these sites would have no property cost advantage. But overall, the
competitor is still benefiting from these lower costs and can use the
money to strengthen its business. Hence it should be part of the Value
Advantage Equation.
Knowledge about locations will enable the competitors to choose
better sites, at least for a few years. We could assess the likely impact
of this based on a typical comparison between better sites and average
sites in the health and beauty company’s current portfolio. We might
give this a value of an average of 5% extra sales per site.
The volume advantages could also be estimated. Let us suggest that
these might amount to a lower cost base of 2% of sales.
106
THE GROWTH GAMBLE
Box 5.4 Net non-tradable contribution in eye care
% of margins
Unique contribution 50
Less tradable portion 25
Less competitor advantages

– Site selection 20
– Scale benefits 2
– Marketing costs 3 (for 5 years)
– Cost per sale 3 (for 5 years)
__
Net non-tradable contribution –3
Brand loyalty is hard to put a number on. However, we could make

a guess at the discounts or extra marketing a new company would need
to give (and for how many years) in order to build up market share. We
might estimate that this would lead to an extra 3% on the cost base for
five years.
Finally, the knowledge of customer preferences will enable com-
petitors to fulfill customer needs more often than a newcomer. This
would reduce their cost per sale. It may also enable the competitors to
successfully sell customers more expensive solutions more often. We
could guess at the impact of this temporary advantage as being a fur-
ther 3% per year over five years.
Putting this all together, we need to judge whether the unique con-
tribution that the competitors are able to make is significantly greater
than, roughly equal to, or significantly less than the non-tradable
contribution the health and beauty company can make. The judgment
is about whether there is a net non-tradable contribution (Box 5.4).
At this point the judgment involves trying to weigh both quantifi-
able and qualitative parts of the different contributions. We should not
pretend that it is easy or that it can be done on a calculator. However,
107
in our experience, managers who understand the business and who

have worked through the details normally come to the same answer. In
this case the net non-tradable contribution is probably about zero. In
other words, there is no strong value advantage logic for the health and
beauty company to enter this business.
ICE CREAM IN EUROPE
If we do the same analysis about whether Mars or Cadbury should

enter the European ice-cream market, we come to a similar conclu-
sion. They clearly had some important contributions to make to this
new business:
❏ Their confectionery brands.

❏ Knowledge of snack products and European markets.
❏ Manufacturing skills.
❏ Branding skills.
❏ Sales and distribution.
❏ Ice-cream experience in North America.
Most of the value that the brands contribute could probably be traded
through licensing or joint-venture agreements, without involving the
companies directly in the risks of the ice-cream industry. The other
contributions, however, would be much harder to trade.
When compared to the main competitors, nevertheless, the other
contributions are not so unique. Both Unilever and Nestlé have direct
knowledge of the ice-cream business in Europe, good branding skills,
established sales and distribution, as well as experience in North
America. Where Mars has an advantage, if any, is in manufacturing.
Most observers would argue that Mars’s manufacturing skills are supe-
rior to those of its consumer product rivals.
A proposal, for example, that Mars enter the ice-cream business
would need to rest on a belief that Mars is much better able to exploit
the value of its brands by entering the business directly, and that its
manufacturing skills are a significant advantage. In addition, Mars
108
THE GROWTH GAMBLE
managers would need to believe that the incumbents, Unilever and

Nestlé, have no significant offsetting advantages.
Since the incumbents have much greater knowledge of consumer
preferences, established relationships with retailers, proprietary distri-
bution units in some retailers, as well as significant economies of scale
in purchasing, manufacturing, and distribution, it is unlikely that Mars
has any positive “net non-tradable contribution.” More probably, if
Mars chose to enter the business through a greenfields operation, it
would have a significant disadvantage.
If, on the other hand, Mars were able to enter the business
through a major acquisition, say Nestlé’s business, the net non-trad-
able contribution could be positive. Mars would be bringing its
brands and its manufacturing expertise to a business that already had
the other advantages. If the premium Mars had to pay for the busi-
ness were less than the value of the non-tradable part of its brands
plus its manufacturing skills, the net non-tradable contribution
would be positive.
The judgment about net non-tradable contribution is therefore
dependent on the precise nature of the proposal. It may be possible to
change the judgment by choosing a way of entering the business that
makes the balance of contributions more positive. For example, if Mars
could do a deal with Nestlé that involved merging Mars’s North
American ice-cream business with Nestlé’s mainly European business,
it might have been possible to avoid paying a premium and hence cre-
ate a large positive net non-tradable contribution.
If a project has a negative net non-tradable contribution, it is not
necessary to look at the remaining terms in the Value Advantage
Equation. The project can be rejected without further analysis. If the
project has a positive net non-tradable contribution, however, it may
still not be supportable. The last term in the equation—cost of learn-
ing the new business—may be sufficiently negative to outweigh the
positive.
109
WHAT IS THE COST OF LEARNING THE NEW BUSINESS?
Insight 2: Managers do not normally assess the likely costs, at

both the operating level and the corporate level, of learning a new
business.
As every manager knows, it takes time to learn a new business. During

this period of learning, the management team makes mistakes that
impose extra costs on the business and misses opportunities that could
bring in extra revenues. The benchmark here is the competition. The
new management team is likely to underperform the competition dur-
ing its learning period. Of course, if the new management team is
experienced in the industry, either because the managers have been
hired from the industry or because the new team is part of an acquisi-
tion, this learning cost is likely to be low. If the new team is from a dif-
ferent industry where different rules of thumb guide the major
decisions, the learning costs can be quite high.
In our sample, which was biased toward situations where the com-
pany had much to learn, we found very few successes where the
company had lots to learn compared to competitors. When the busi-
ness is new to the world every competitor has lots to learn, hence the
relative position is more even (Figure 5.5).
In addition to the learning that needs to take place at the operating
level, there is also learning required at the corporate level. The new
business will have a reporting relationship with its parent company and
will share resources or functions with some other businesses. A new
eye-care unit in the health and beauty retailer might report to the busi-
ness development director, who in turn reports to the CEO. In Mars
or Cadbury, a new ice-cream business might report to the head of
Europe, who in turn reports to the worldwide CEO.
These reporting relationships are influential: the management team
in the new business is significantly affected by the signals, guidance,
and instructions that it gets from its corporate masters. Beyond these
reporting relationships, the management team is also influenced by a
range of corporate functions that may control human resource policies,
110
THE GROWTH GAMBLE
Figure 5.5 Levels of business unit learning
accounting policies, IT choices, and other factors. In the case of Mars,

an ice-cream business would have relationships with support functions
such as HR, the European research and development function, and the
country-based sales and marketing organizations. Each of these func-
tions has established ways of working that may or may not suit the new
ice-cream business.
The question that is being asked at this point in the Value
Advantage Equation is about the impact of these influences. Do the
managers higher up in the hierarchy and their supporting functions
understand the new business well enough to give it good guidance and
advice, or do they need to learn which parts of their normal influences
are appropriate to this new entity and which parts are poisonous? Do
the shared resources, such as sales or manufacturing, understand the
new business well enough to be a help rather than a hindrance?
Few managers attempt to assess the costs of learning at both the
operating level and the corporate level. None, in our experience, puts
these costs into the financial analysis. Figure 5.6 overleaf captures these
two dimensions of learning. The learning needed at the operating level
of the new business is shown on the vertical axis. The learning needed
111
Figure 5.6 The learning challenge
by the corporate managers is shown on the horizontal axis. The more

learning required, the greater will be the learning costs.
The difficulty with learning costs is to judge how big they are. The
judgment on the vertical axis is about the degree to which the business
model of the new business and the types of relationships managers
need to have with customers, suppliers, employees, unions, regulators,
and other stakeholders are similar to managers’ previous experience.
What is more, the accumulated experiences of managers generally get
built into a company culture, a set of beliefs and behaviors that are
often taken for granted by the people involved. This culture pervades
the way of doing business, the way of dealing with outside relationships
as well as with one another in the company. If the business model or
the relationships, or the required beliefs, behaviors, and organizational
structures and processes, are different for the new business, the learn-
ing costs will be high. If the business model, relationships, and cultural
requirements are familiar, the learning costs will be low.
Unfortunately we do not have a clear calibration for what “high”
and “low” mean. We have seen managers with inappropriate
112
THE GROWTH GAMBLE
experience make huge mistakes, resulting in the complete failure of an

otherwise promising new business. When WHSmith, a stationery and
newsagent company, entered the do-it-yourself (DIY) sector, managers
were transferred from the newsagent business into the DIY business.
The newsagent business was a mature, dominant retailing format
where good performance was easy and improvements were about store
ambience, systems enhancements, and product selection. The DIY
business was growing fast. Success depended on opening new sites
faster than competitors and aggressive branding and pricing in order to
get a market leadership position that would give volume advantages.
Not surprisingly, the managers from the newsagent business were slow
to change, spent too much energy on tinkering with the format, and
missed the opportunity to get a leadership position. Do-It-All, the
business they were managing, never made acceptable profits and was
eventually given away after incurring a string of heavy losses.
If the management team is from the industry, particularly if it is part
of an acquisition, the learning costs at the operating level may be zero.
More normally they are somewhat negative (at least 5% impact on
value) and frequently significantly negative (30% or more impact on
value).
In the case of the eye-care and ice-cream examples, it is tempting to
think that the learning costs would be low. The health and beauty man-
agers are already experienced retailers, and managers in Mars or
Cadbury are experienced in consumer snacks. However, in both exam-
ples managers will discover things they need to learn. For example,
when Mars entered the business, it launched test products at the time
of the two hottest summers in Europe for more than 70 years. Without
data going back 30 years and without experience of how to read the
market, Mars managers were unsure how much of the market increase
was due to their new product offerings versus the summer heat. As a
result, they overestimated the volumes they would get and conse-
quently built a factory that was significantly too big. This error alone
had a negative impact on the value of their new business that was at
least 10% and probably exceeded 30%. Since it is difficult to predict
this kind of learning cost, we believe it is wise to presume a somewhat
113
negative learning cost (i.e., 10%), unless the management team has
worked for a few years in an identical product/market area.
The judgment on the horizontal axis is also difficult. How far will
the parent managers, functional heads and shared resource leaders
need to modify their normal behavior and policies in order to give the
new business good guidance and influence? If the business can be man-
aged in an identical manner to the other businesses in the portfolio,
then learning costs will be low. If the new business requires the parent
to learn new behaviors that are compatible to its existing habits and
instincts, the learning costs will be significant but manageable. If the
new business needs to be managed in a different way from the existing
businesses, requiring less compatible behaviors and organizational
approaches, the learning costs are likely to be high: over 30%.
For Mars, one of the normal rules of thumb when entering a new
category was to build a factory larger than was immediately needed. In
the past, this had proved beneficial for two reasons. First, the large
empty factory provided the best kind of incentive to the sales team to
work hard. As one of the Mars brothers is reputed to have said:
Build a factory larger than you need and it will cause the sales force to
work twice as hard to fill it.
The second benefit comes from economies of scale and market share.
The greater the volume relative to competitors, the more competitive
the operation will be.
Unfortunately, this corporate influence was instrumental in encour-
aging the European managers to build a large ice-cream factory. But,
partly because of the misleading market data and partly because the
sales forces did not feel responsible for the factory—since each sales
force reported to a separate country and sold confectionery and ice
cream and some other products as well—the factory turned out to be
much too large. The influence of the parent company had been
negative.
Mars’s human resource policies were probably also negative. The
ice-cream job was allocated a grade that made sense within the com-
114
THE GROWTH GAMBLE
pany’s broader grading system. However, it made it hard to recruit a

manager with the political weight, market knowledge, and strategic
skills needed to lead the ice-cream business effectively. Again, the
habits of the parent company had a negative influence on the success
of the business.
In our experience, stories like these are the norm rather than the
exception. When Shell entered the aluminum business, its habits
encouraged the aluminum management team to pursue a vertical inte-
gration strategy. This had proved a successful strategy in oil, and hence
there was some, often unspoken, pressure to try it in aluminum. The
result was bad. Shell discovered something that most managers in the
industry already knew: that vertical integration does not create value in
aluminum.
When Xerox tried to translate its remarkable R&D efforts into a
new office-automation business, comprising the Star workstation and
the Ethernet, a wordprocessor and a laser printer, the inclination of
corporate managers was to require the same product review and
involvement in decision making they were used to in the copier busi-
ness. They had difficulty understanding a business model that required
rapid innovation cycles, indirect sales, lower margins, and third-party
software. When they hired an outside team and gave them free rein,
this team alienated the rest of the company. Xerox ended up abandon-
ing the new business and concentrating on the Japanese challenge in its
copier business.
The Mars, Shell, and Xerox examples are common. In our database
of successes, which was biased toward situations unfamiliar to the par-
ent company, we found few successes where the parent company man-
agers had lots to learn (Figure 5.7 overleaf).
So how can managers predict whether the impact of the parent
company is going to be significantly negative? There are four factors
to consider:
❏ The similarity of the business model of the new business to that of the core
businesses. If the model is similar, the parent influence is less likely to
be negative.
115
Figure 5.7 Levels of parental learning
❏ The similarity of the needs of the new business to that of the core businesses.
If the new business requires similar cultural beliefs and behaviors
and needs similar kinds of support from its parent, the influence is
less likely to be negative.
❏ The commitment and experience of the relevant line managers. If the line
managers to whom the business reports are committed to the new
business and have had experience in this new industry, there is less
chance of a negative influence.
❏ The demands of other commitments. If the line managers are highly
committed to this new business and have few other commitments,
there is less chance of negative influence.
The managerial temptation is to be positive: to conclude that the par-

ent company influence will be neutral or positive. Our experience sug-
gests that managers should presume that the parent company influence
will be a negative 10%, unless there are strong reasons to suppose
otherwise. The reason for taking this negative starting point is that,
where new businesses are concerned, we have seen more examples of
negative influence than positive.
116
THE GROWTH GAMBLE
While we recognize the difficulty of putting precise numbers on the

overall cost of learning the new business, the biggest danger is to
ignore this term of the equation altogether: to assume that managers
will be effective in the new area and that the learning costs will be zero.
A safer alternative is to assume that the learning costs will be signifi-
cant (10% at the operating level and 10% at the parent company level),
unless there are strong reasons to suggest that they should be smaller
or greater.
HOW ATTRACTIVE IS THE PROFIT

POOL?
Insight 3: Managers focus too much attention on the potential of

the marketplace they are thinking of entering. Strategy analysis
tells us that, in normal situations, a company will earn above-
average returns only when it has competitive advantage. Hence
analysis of markets should focus on identifying extreme situations
that are either so good that even a competitor with disadvantages
can earn a good return, or so bad that even an advantaged player
may earn less than the cost of capital.
The attractiveness of the profit pool is the second Traffic Light. The
profit pool question is about whether the market is a “rare game”
(green), a golden opportunity that is likely to give good returns to most
competitors who invest today, or a “dog” (red), a rotten market that is
either too small, too uncertain, or too competitive, leading to low
returns for most players to the point where even the market leader will
have a hard time. All other situations, markets that are neither dogs nor
rare games, are “possibles” (yellow). If the company has some advan-
tage and executes well, it should make good money.
To help managers identify dogs, possibles, and rare games, we
have defined five criteria that need to be assessed (see also Box 5.5
overleaf):
117
Box 5.5 The potential of the profit pool
Business model potential for high margins

– Value perceived by customers relative to cost of production
– Break-even volume as percentage of market
– Revenue model
Industry structure potential for high margins

– Power of customers
– Power of suppliers
– Competitive rivalry
– Threat from new entrants
– Threat from substitutes
– Growth
Opportunity for us to be a leader in this market
Cost of trying relative to size of profit pool

– Taking account of time to commercialization
Business model vulnerability

– Number of enablers
– Sensitivity to key variables
❏ Business model potential for high margins.

❏ Industry structure potential for high margins.
❏ Opportunity for us to be a leader in this market.
❏ Cost of trying relative to size of the profit pool.
❏ Business model vulnerability.
The first three questions address the potential for earning a positive
spread over the cost of capital. Is the value being offered so important
that the customer is likely to pay significantly more than the costs of
providing the service? Is the industry one where big margins will not
118
THE GROWTH GAMBLE
be competed away or bargained away? Is there an opportunity to

become the leader and hence earn margins higher than the average?
The next two questions address two other factors critical to the
judgment about the attractiveness of the profit pool. Is the ultimate
prize—the value of the business once it succeeds—worth the risk, the
cash that will need to be put at risk in order to enter the market? How
much is the risk under our control: is the profit in the pool vulnerable
to the whims of other players, volatility in key variables, or movements
in the cycle?
Thus, rare games are markets where the profit potential is so good
that almost any competitor entering the business at the right time is
likely to make a positive return on its investment. In rare games, the
business model gives room for high margins, the industry structure will
allow high margins to be earned, and there is an opportunity for the
company to become the leader. In addition, the company can enter the
market without a big cash exposure and the risks are mainly ones under
its control. In other words, all the sub-points are favorable.
The market for internet service providers in the UK in the late
1990s is an example of a rare game. The business model suggested that
high margins were possible. The value delivered was significant. Yet
the cost of servicing each incremental customer was very small. The
volume needed to break even was small compared to the size of the
market. The revenue model, which depended on discounts from the
telecoms companies and advertising on the portal, made it possible to
capture a share of the value being created. In addition, the industry
structure was favorable. Customers had little bargaining power. The
telecoms providers were more interested in expanding capacity than
exploiting their bargaining power. There were no obvious substitutes.
The only negative factors were the ease of entry and hence competi-
tion between suppliers. Because of the market growth rates and the
revenue model, which allowed companies to offer a free service, rivalry
was muted. Another positive factor was the availability of the leader-
ship slot, common in new markets. Even the cost of entering the mar-
ket was small compared to the potential profit pool. The only negative
factor was uncertainty about future regulation.
119
In this rare game situation, many new businesses started up, all of
them earning reasonable returns. The leader turned out to be
Freeserve, launched by Dixons, a retailer of electrical products. Within
three years Freeserve was valued at over £2 billion.
In contrast, dogs are businesses where a weak business model, an
unattractive industry structure, and a lack of opportunity for leadership
combine with high levels of vulnerability and high risks relative to
reward.
The full-service airline business in the US has been a dog market
for a number of years. While the business model is not inherently
unattractive, the industry structure is bad. The main problem is the
substitute: low-cost airlines. Full-service airlines have had to price
down to compete with the low-cost companies. In addition, the high
fixed costs of the industry have encouraged the full-service companies
to compete vigorously with each other as well. When this is combined
with the difficulty any new airline would have in becoming the indus-
try leader and the high costs of getting into the industry compared to
the size of the profit pool, it is not surprising that no new full-service
airlines have been launched in the last 20 years.
In summary, the five elements of the profit pool criteria (business
model, industry structure, leadership opportunity, risk/reward ratio,
and vulnerability) are not particularly special or unique. The clever
part is the insight that managers only need to make a green, yellow, or
red assessment. Moreover, since 80–90% of markets are yellow, mean-
ing that they offer reasonable returns for competitors with advantage,
managers only need to identify the few outliers. The five elements help
managers identify these outliers.
WHAT IS THE POTENTIAL OF THE BUSINESS MODEL FOR

HIGH MARGINS?
The business model has the potential for high margins when three
conditions exist. First, the value provided to the customer needs to be
high relative to the variable (or per-unit) cost of production. A pill sold
by a pharmaceutical company is a classic example. The manufacturing
120
THE GROWTH GAMBLE
cost of the pill may be a few cents, but the value to the customer of bet-
ter health may be in the thousands of dollars. Similarly, a micro-
processor chip made by Intel is relatively inexpensive to manufacture
compared with the valuable function it performs in a PC. In both cases,
the value is high relative to the cost. The eye-care business is also one
where the value to the customer is high relative to the cost of provid-
ing the care. The ice-cream business is more balanced. The value is
greater than the cost, but not hugely so. This makes high margins
harder to earn.
The second factor affecting the potential for high margins is the
break-even volume. In investment-intensive businesses, such as steel,
where the cost of capacity can amount to billions of dollars, a company
may need a large market share to pay for the capital investment. Other
businesses such as retailing are less capital intensive, but have high lev-
els of fixed operational costs once a given service level and geographic
coverage have been set. These businesses often also require a big share
of the market to reach break even.
A pharmaceutical business will require sales of hundreds of millions
of dollars to pay back its research costs. A chip manufacturer requires
even higher volumes to achieve payback on a new silicon fabrication
plant. An eye-care retailer needs to attract a significant share of local
trade in order to break even. In ice cream, however, capital intensity is
modest, and the break even is low. A small factory may need to oper-
ate at above 70% capacity to make good returns, but the factory will
only serve a small proportion of the market. Generally, the larger the
up-front fixed costs in plant, in service infrastructure, or in developing
the product, the bigger the percentage of the market needed to break
even.
The third factor is the potential for transforming high customer
value into high margins. Businesses selling a physical product are gen-
erally able to price their product to reflect the perceived value, depend-
ing of course on the intensity of competition (dealt with further
below). In certain instances, social considerations and regulation may
place constraints on value-based pricing for products like pharmaceu-
ticals or utilities. When the product carries content, like software or
121
music on a CD, value-based pricing can also be difficult. Unless good

safeguards can be found, copying or pirating makes it hard to maintain
high prices. Equally, businesses selling a service often find it hard to
capture the value they create unless they have a business model based
on usage or on a subscription. If neither is feasible, as in many types of
internet services, advertising may be the only way to derive revenue
from the value provided to customers. The revenue model obviously
has a critical feedback effect on volume, and therefore on break even.
These three factors affect the size of the available margin and hence
the potential profitability. The most unfavorable factor dominates the
assessment. In ice cream, the most unfavorable factor is the value-to-
cost ratio. The conclusion for ice cream is, therefore, that the business
model potential for high margins is average: neither very good nor very
bad. In eye care the most unfavorable factor is the market share
required in most locations to justify a retail unit. Again, the conclusion
is average: eye care does not offer opportunities for exceptional
margins.
WHAT IS THE POTENTIAL OF THE INDUSTRY

STRUCTURE FOR HIGH MARGINS?
The economics of the business model tells us the potential for profit,
but it is the structure of the industry that determines what profit is
actually earned. Michael Porter’s five forces model is the ideal tool for
assessing the structure (Figure 5.8). This model identifies five forces—
customer bargaining power, supplier bargaining power, threat of new
entrants, threat from substitutes, and competitive rivalry—as being the
determinants of actual profitability. In other words, the companies
competing in an industry sector, such as pharmaceuticals, eye care, or
ice cream, can only earn the profits that their customers, suppliers,
rivals, potential entrants, and substitutes allow.
Powerful customers can have a negative impact on profitability.
They normally bargain down the price in order to capture as much
margin as possible for themselves. Automotive companies are well
122
THE GROWTH GAMBLE
known as very effective bargainers for low parts prices. Powerful sup-
pliers can also have a negative impact. They also bargain on prices:
they push their prices up to suck margin into their business and away
from their customers. Intel and Microsoft have been able to exact high
prices from PC manufacturers, contributing to the low profits earned
by PC companies.
A third force is the intensity of rivalry between competitors.
Competitors can be so intent on winning share from each other or
competing for leadership that they cut prices, leaving little margin.
Home electronics suppliers, such as Matsushita and Philips, have been
notorious examples of this behavior. For decades, they have tried to
gain superior global market share and dominate industry standards by
launching products at attractive prices. The result has been low prof-
itability for most competitors.
The fourth force comes from companies outside the industry. They
may enter the market if margins look attractive. This threat from pos-
sible new entrants causes existing competitors to keep margins down in
order to discourage other companies from coming into the market.
This is common in some high-tech industries, such as semiconductors,
where prices are frequently kept low to discourage new capacity.
Figure 5.8 Five forces analysis
123
The fifth force comes from substitute products. Companies often

keep their prices down in order to discourage customers from choos-
ing substitute products. Local telecom companies in the US have
reduced prices to fend off new entrants and substitute products, like
voice over internet.
Any of these five forces can be the cause of low profitability.
Companies only earn the full profit potential of their business models
when customers and suppliers do not bargain on price, entrants are no
threat to prices, there are no price-based substitutes, and competition
is muted.
These favorable conditions exist in eye care and pharmaceuticals,
but not in ice cream or auto manufacturing. In ice cream in Europe
there are few problems with customers, suppliers, new entrants, or
substitutes. But the competitors, Unilever and Nestlé, are unlikely to
take a new entrant lying down. Rivalry, which is normally not intense,
would be likely to increase if a new company entered the business. This
would reduce margins for all competitors. In other words, Mars or
Cadbury would be likely to earn less than average margins because of
competitor reactions. In eye care, a new entrant is likely to earn
average margins because the five forces are reasonably favorable.
When assessing a new business the growth rate of the market is nor-
mally considered to be vital. Managers place a value on growth simply
because growth in revenues and profits is one of the main reasons for
exploring new businesses. A marketplace that is growing fast is, there-
fore, a prize in its own right.
From the perspective of profit pool potential, however, growth is
good only so far as it helps raise profitability. It is not valuable in its
own right. Growth often affects profitability because it lessens negative
forces in Porter’s five forces model. Growth makes the customers less
demanding because in high-growth markets customers are more inter-
ested in the reliability of supply than in the last cent or two of price.
Suppliers are less likely to be greedy for margin if they see steady
growth. Competitors are more concerned with expanding than fight-
ing for share. Any new entrants that come into the business will be less
disruptive because they will be less likely to create volume declines for
124
THE GROWTH GAMBLE
the existing players. Hence high growth—above 10% volume

growth—is normally associated with high profitability.
But growth can be bad news when an industry becomes overhyped.
High-growth markets can attract too many competitors. Since most
companies are looking for growth, every growth market will be a
potential new business opportunity for a number of companies.
Frequently, these markets are oversupplied with competitors as each
tries to establish a foothold in what they expect to be a promising
future. The cellular telephone industry was an example of this in the
1990s, as was the business of fiberoptic communications. The results
are high growth and low profits as the threat of new entrants increases
and industry rivalry becomes high.
Both eye care and ice cream are growing markets, but neither is
growing so fast that it is undersupplied or has become attractive to a
large number of new entrants. For these two examples the growth fac-
tor is neutral.
WHAT IS OUR OPPORTUNITY TO BE A LEADER?
The potential for the company to achieve a leadership position is

important to long-term profitability. Leaders earn greater margins
than followers. Leaders are in a position to influence the rules of the
game, set standards of performance, and reap scale economies.
In Profit from the Core, Chris Zook and James Allen provide some
data assembled by their consulting company, Bain & Co. Based on a
sample of 185 companies in 33 industries, they conclude that there is a
linear relationship between return on capital and leadership. Strong
leaders earn 25%, weak leaders 22%, companies at parity earn 14%,
strong followers 9%, and weak followers 4%. In the long term leader-
ship is important. Followers have inefficient economics and are more
vulnerable to changes in the business that are imposed or readily
embraced by large players.
Unfortunately, the path to leadership is sometimes blocked off. The
market may have an established leader or a number of large competi-
tors that are unlikely to be up for sale.
125
If the industry is new to the world or highly fragmented, the leader-

ship position will be open and the opportunity becomes more attrac-
tive. Industries requiring high up-front investments in technology,
capacity, or infrastructure provide more of an assurance of potential
leadership to the first or most-endowed mover than industries where
the cost of entry remains low for second or third movers. When the
first mover is able to set an industry standard, which will then be used
by others, a leadership position is almost assured.
More mature or highly fragmented industries may offer the oppor-
tunity of seizing leadership in two ways. One is to buy several existing
competitors and consolidate the industry. Another is to upset the sta-
tus quo with a new business model, technology, or marketing
approach. Blockbuster was successful with the latter strategy in the US
video rental business and Wal-Mart in the variety retailing business.
In the European ice-cream business the opportunity for leadership
is blocked. In Europe, ice cream is dominated by Unilever, with Nestlé
a distant number two. However successful Mars and Cadbury might
be, they could never hope to wrest leadership from Unilever. In eye
care there are dominant competitors, but a new entrant from the
health and beauty business might expect some of the eye-care com-
petitors to come up for sale, providing an opportunity to become the
leader in the longer term.
WHAT IS THE COST OF TRYING RELATIVE TO THE SIZE

OF THE PROFIT POOL?
Entering a new business requires investment. Until the business has

established itself, this investment is at risk. The assessment we are sug-
gesting, therefore, is a ratio of the size of this investment at risk (cost
of trying) to the size of the profit pool in this new market.
Frequently the amount of money that has to be put at risk is large.
This is especially true for new-to-the-world businesses that require
many years of technical development and test marketing before the
business case can be proven. An example is the domestic combined
heat and power project at BG Group. The objective was to create a
126
THE GROWTH GAMBLE
small gas generator that can be sold to households to provide heat and
electricity. After more than 10 years of research the product is still not
on the market. Another example is the emergence of electronic deto-
nators for explosives. First developed in the late 1980s, the technology
has taken more than 10 years to become commercial.
The assessment that managers have to make is whether the market
is large relative to the cost of proving that the business can be profitable
for the company. If the total investment needed before profitability has
been proven is $10 million, the market would need to be at least $100
million to be attractive. This ten-to-one rule of thumb is based on some
simple calculations. If we assume the company gets 20% of the final
market and earns a 10% margin, there will be a margin of $2 million per
annum available to pay for the start-up costs. It will, therefore, take five
years to pay back the start-up costs of $10 million—hence our rule of
thumb that the market needs to be at least 10 times the risk investment.
Any market size less than 10 times the risk investment is a potential dog.
A really attractive market is 50 times the size of the risk investment (i.e.,
the start-up costs can be recovered from one year’s profit assuming a
20% share). If the share ambitions are much lower than 20%, then the
size-to-investment ratio needs to be much larger.
For the ice-cream and eye-care decisions, both markets are more
than 50 times the risk investment needed to prove the opportunity.
HOW VULNERABLE IS THE BUSINESS MODEL?
Some business models depend on the cooperation of partners or other

stakeholders for their success. For example, when Apple launches a
new computer platform, its success is dependent on gaining the sup-
port of software companies to produce applications that will run on its
platform. When Intel launches a new microprocessor, it is dependent
on the decisions of a few computer manufacturers. With a large share
of the market, Intel’s business model is secure. But any new entrant is
vulnerable to decisions taken by a few players.
The same is true of businesses in regulated industries where
changes in regulation may protect or expose incumbent suppliers.
127
Often the regulatory framework is quite stable and does not make the
business model vulnerable. But in industries like telecoms or health
care, uncertainties about regulation can create significant vulnerability.
A third cause of vulnerability comes from joint venture partners.
Business models that depend on the commitment of joint venture
partners or even suppliers of some key ingredients can be vulnerable to
changes in the strategies of these companies.
A further source of vulnerability can come from competitors, who
may react in unforeseen ways that limit the capability of a new entrant.
Competitors in the electronics industry, for example, have defended
their positions by locking in customers through the use of standards
and cross-business or multicountry deals, which make market access
difficult for a newcomer.
Finally, business models can be vulnerable to movements in critical
variables—exchange rates, commodity prices, or prices of substitute
products.
The telecoms industry provides an example of high vulnerability. A
telecoms operator intending to roll out a new 3G network will depend
not only on regulation to permit the construction of the network but
also on investments made by suppliers of handheld devices and
providers of services. Failure to get support from these bodies will
leave the telecoms company without a market for its network.
Both the eye-care and ice-cream businesses appear, on first analysis,
to have low vulnerability. They do not depend on partners, suppliers,
or regulation, and they are not overly exposed to movements in raw
material prices or exchange rates. Deeper analysis, however, would
expose a key vulnerability for any new entrant in the European ice-
cream business. Unilever and Nestlé own some freezer compartments
in retail outlets and will be likely to try to prevent any new entrant
from placing its products in these compartments. In a worst-case sce-
nario, this could prevent Mars and Cadbury from distributing their
products through some retailers.
128
THE GROWTH GAMBLE
CALIBRATION
We have not tried to calibrate each of these variables in the profit pool
potential or work them into some equation. The reason is that the
judgment required—dog (red), possible (yellow), rare game (green)—
does not demand accurate calibration.
A dog market is one where:
❏ the economics of the business model are poor, meaning that the
value to the customer is only slightly greater than the cost of pro-
duction and the break-even volume demands at least a 50% market
share; or
❏ the industry structure, even with the benefits of growth, is such that
few if any competitors will cover their cost of capital; or
❏ the cost of proving the idea is more than one tenth of the market
size; or
❏ the business model is dependent on partners, stakeholders, or vari-
ables that are uncertain and could be disastrously negative.
These markets should be avoided regardless of the outcome of the

other parts of the Traffic Lights.
A rare-game market is one where all the criteria are favorable, where:
❏ the value to the customer is more than twice the variable costs of
production and the break-even point requires a market share of less
than 5%; and
❏ the five forces are all favorable and growth is greater than 10%; and
❏ there is an opportunity to become the market leader; and
❏ the market size is more than 50 times the investment needed to test
the opportunity; and
❏ the business model is not vulnerable to any major uncertainties.
In these markets, companies can contemplate investing even if they do

not have a clear advantage and even if the impact on the existing
businesses is slightly negative.
129
Managers have found it relatively easy to use this tool. Around 90%
of markets turn out to be yellow. Hence the analysis is normally
focused on making judgments at the boundaries. A few markets will be
borderline rare games and a few will be borderline dogs. Rather than
agonize over which side of the border to put a market, it is often bet-
ter to acknowledge that it is a borderline case and move on to the next
part of the Traffic Lights.
For both of the two case examples—European ice cream and eye
care—the assessment is yellow. The business models have average
potential for high margins. The five forces are worrying in European
ice cream, but unlikely to warrant a red light: Unilever and Nestlé will
compete hard but are unlikely to destroy pricing in the industry.
Neither market offers easy access to market leadership. Neither is small
relative to the cost of entry. Neither is unusually vulnerable. On bal-
ance, the European ice-cream market appears to have a less attractive
profit pool than eye care, but both fall in the middle, yellow category.
One final point: A thorough analysis of all the factors under profit
pool potential requires a great deal of data. This is rarely available. Our
advice is to start by assuming that the profit pool is “yellow.” Then
scan down the five elements to see if any appears to suggest that a “red”
or “green” assessment might be more appropriate. If a red or green
assessment seems possible, more detailed analysis will be necessary. In
European ice cream the only factors that might signal an assessment
other than yellow are the competitive rivalry element in the five forces
and the fact that leadership is blocked off. Closer analysis of these two
factors would then be necessary.
DO WE HAVE STRONG LEADERS AND

EFFECTIVE SPONSORS?
Insight 4: Managers do not give sufficient attention to the issue

of who is going to run a new business and, particularly, who the
new business is going to report to.
130
THE GROWTH GAMBLE
The third Traffic Light is about people. Our research, evidence from
the venture capital industry, and pure common sense all point to the
importance of people. In our successful cases we were struck by the
unique people involved, the often chancy way in which they had come
together, and the special experiences that influenced their thinking. In
the venture capital and private equity industry a well-worn phrase is
that there are only three factors for success: management, manage-
ment, and management. And common sense tells us not to enter a team
in a sporting competition and expect to win unless we have some
exceptional players when compared to the opposition. People are
important.
Yet new business projects are frequently launched, with the help of
managers in a business development function and supported by con-
sultants, on the presumption that the operating management team can
be recruited later. If the activity is familiar and the company has a pool
of capable managers ready to step up to the challenge, this approach is
reasonable. However, if the activity is less familiar, involving a differ-
ent business model, it will be hard to find the necessary management
talent.
The questions that need to be addressed under this Traffic Light are
(see Box 5.6 overleaf):
❏ Do the leaders of the new business (the unit head and his or her
team) have the passionate commitment, personal insights, entrepre-
neurial flexibility, execution skills, and influence with the parent
that will enable them to overcome the inevitable setbacks, skepti-
cism, and roadblocks, and win in the marketplace?
❏ Does the new business have a sponsor who will provide a compati-
ble home for it in the portfolio, exercise effective oversight, protect
it from negative influences, and support it through setbacks?
HOW GOOD ARE THE UNIT LEADERS?
We need to make judgments about some important qualities in the

leaders of any new business. First, we need to assess passion and
131
Box 5.6 Leadership/sponsorship quality
Relative quality of MD/leadership team of the unit

– Commitment
– Personal insights
– Entrepreneurial flexibility
– Execution skills
– Influence with the parent
Status of sponsor within main parent

– Significant line manager
• Channel resources to the new business, especially during setbacks
• Persuade other businesses and services to give help
• Protect the business from overenthusiastic functions
– Compatible home and effective oversight
commitment. Do the leaders of this new business feel passionate about

it? Are they going to struggle hard when the going gets tough? Are
they going to be able to remain positive when faced with setbacks?
Second, we need to understand the leaders’ personal insights. Do
the leaders have personal insights about this business based on
experiences they have had in other jobs or in the early pursuit of this
project? Managers who have personally serviced the first few clients,
designed the first few products, assembled and delivered the first
orders have often learnt some important lessons and developed advan-
tages over those who have only ever seen the business from the top
down.
This combination of commitment and personal insight is necessary
when managers are faced with setbacks, skepticism, and roadblocks.
The personal belief that “I know there is a pony in here somewhere”
will provide the motivation needed. As one of the two founders of
Homeserve, a home plumbing and support business that sells annual
insurance, said:
132
THE GROWTH GAMBLE
We would never have succeeded without Richard [his partner]. He

remained convinced that we would succeed, even when all the data was
telling us that we were failing.
This faith was based on some personal insights that the two ex-Procter
& Gamble managers had developed in their early attempts to get into
the plumbing business. They had learnt the following:
❏ Finding an honest plumber is difficult, especially in an emergency.

❏ Customers are concerned about overcharging, even though they are pre-
pared to pay a premium for a good job done in a hurry.
❏ It ought to be possible to provide a plumbing service like the cover you get
for your boiler: an annual insurance with guaranteed service levels.
❏ Direct mail is an excellent marketing tool because it is possible to pinpoint
the household that is likely to be a customer.
The judgment about personal insights is not all about experience. In

fact, long years of experience in a sector can create locked-in mindsets.3
A management team that has not been programmed by history may be
much better able to read the signs about impending changes than the
incumbents.
The third quality is entrepreneurial flexibility. One certainty is that the
original business plan will need to change and the market and competitor
landscape will develop in unexpected ways. As Barry Gibbons, ex-CEO of
Burger King, explained: “Your business plan contains the one and only
scenario that is guaranteed not to happen.”4 In the face of this reality, the
team in charge of the new business needs to have the entrepreneurial
skills and courage to adjust its plans when circumstances change.
The fourth quality is execution skills. New businesses involve set-
ting up new operations, systems, and organizations. We need to be
confident that the management team has the skills to execute these
business-building tasks. This is partly a matter of functional expertise
and partly about execution orientation.
The final quality is influence in the parent company. For some new
businesses, such as the ice-cream market example, links with the core
133
businesses are critical for success. The new business is likely to depend
on other businesses to sell or manufacture or market its product. But
even for new businesses with minimal links to the core businesses, as
would be the case in the eye-care example, influence within the parent
company is necessary to get funding and gain the freedom to break the
corporate rules.
The overall assessment of the unit’s management team, therefore, is
about some specifics, such as insights and influence, but it is also about
the overall quality of the management team compared to others. If the
different management teams in the industry were put up for auction,
would our team sell for a significantly higher or lower price than that
of the lead competitor?
If the team is led by top-quality managers, with some track record
of taking advantage of new developments and reacting swiftly to new
events, some experience or insight relevant to this business, and some
passion for success, we can grade unit leadership as green. Sir Peter
Davis, then CEO of Prudential, one of the UK’s leading insurance
companies, hired Mike Harris to run Egg, an internet banking venture.
Mike Harris had previously started up First Direct, the UK’s first tele-
phone banking operation. He had been involved in converting First
Direct to an internet bank as the technology changed. He had also had
other successful new business experiences. He and the team he
gathered around him were head and shoulders above the management
teams of the other internet banks that were starting at that time.
When Royal Bank of Scotland invested in Direct Line, Britain’s
first direct insurance company, the team led by Peter Wood consisted
of two managers with especially strong IT skills, as we saw in Chapter
3. The combination of the entrepreneurial skills of Peter Wood and
the IT skills of his colleagues Mike Flaherty and Roy Haveland, all of
whom had worked together and innovated together for their previ-
ous employer, Alexander Howden, were far superior to those of other
direct insurance companies set up around that time. As IT capabili-
ties developed and changes happened in the marketplace, the Direct
Line team was quickest to embrace the opportunities and adjust its
business model. In fact, the main rival to Direct Line turned out to
134
THE GROWTH GAMBLE
be Churchill, a company set up by one of the original Direct Line

team.
Stories of personal insight, entrepreneurial drive, unique skills, and
leadership courage are common in texts on innovation and new busi-
nesses. Gary Hamel tells the tale of IBM’s internet success in a remark-
able Harvard Business Review article.5 He focuses on two characters:
David Grossman, “a typical self-absorbed programmer and mid level
IBMer stationed at Cornell University’s Theory Center, a nondescript
building hidden away in the south east corner of the engineering
quad,” and John Patrick, “a career IBMer and life long gadget freak,
who had been head of marketing for the hugely successful ThinkPad
laptop and was working in corporate strategy scouting for his next big
project.” By chance the two link up and become the spark that “built a
bonfire under IBM’s rather broad behind.”
More normally, however, the team chosen to set up a new business
has no claim to be the most capable and entrepreneurial in the mar-
ketplace. Frequently it consists of a technical person who has been
working in the parent company’s research function, a commercial per-
son with analytical skills who has been involved in assessing the viabil-
ity of the project, and a CEO recruited from outside. The chances of
this combination being the winning team is low. Hence we suggest that
managers start by presuming that this part of the leadership/sponsor-
ship Traffic Light will be a reddish yellow unless there is strong evi-
dence to suggest otherwise.
The only exception to this negative presumption is when the new
business is new to all competitors. When Dixons launched Freeserve,
a UK internet service provider, the business was new to all competi-
tors. Dixons had no reason to believe that its managers were clearly the
best in the industry, but equally it had no reason to suppose that they
were inferior. When the business is new to all competitors the starting
presumption should be one of neutrality and a yellow score awarded
unless there is clear evidence either way.
In our sample of successes, over two-thirds involved management
teams who, with hindsight, were unusually strong. In other words, over
two-thirds of the successes would have scored a green light for
135
leadership/sponsorship. The venture capital mantra of “management,

management, and management” appears to be borne out by the data.
HOW EFFECTIVE ARE THE SPONSORS?
Sponsors are important because they can make up for qualities that are
lacking in the management team of the new business, because they
have authority over the new business, and because they can help the
new business get the positives from and avoid the negatives in the
broader parent company. The sponsor should have the following
qualities:
❏ He or she should be a manager in the main reporting line with some

significant businesses reporting to him or her. The CEO of the par-
ent company or the head of a division or a region is a good sponsor.
The head of business development, the head of a corporate ventur-
ing unit, or the head of research would be a less good sponsor.
❏ The sponsor needs power to be effective at three critical tasks:
channeling resources to the new business, especially during times
when resources are scarce; persuading other businesses, corporate
functions, or shared resources to give help to the new business; and
protecting the business from inappropriate influences from
corporate functions or other parts of the organization. A sponsor
with significant power in the hierarchy and with significant
discretion over resources is ideal. He or she can channel resources
to the new business when the official supply dries up. He or she can
provide hidden subsidies to the new business when financial per-
formance is weak. Most importantly, he or she can support the new
business when doubters, skeptics, and setbacks undermine the orga-
nization’s commitment. Without a strong sponsor or the likelihood
of one, it is often unwise to embark on a new business.
❏ In addition to having control of some resources and having power
and influence in the structure, the sponsor should also provide
effective oversight and an understanding home for the new
business.
136
THE GROWTH GAMBLE
❏ The sponsor needs some personal knowledge of the new business or

the time and inclination to learn fast.
❏ He or she must be able to spend the time to challenge and coach the
business and share in some of the critical decisions that need to be
taken along the way.
❏ The sponsor should be influential in the development of the new
business. The more the sponsor has the ability, motivation, and real
understanding to exert that positive influence, the higher the
chances of success. That is why a CEO, or a line manager, is only a
good sponsor when the new business is an important part of that
manager’s strategy.
Companies often set up business development divisions and corporate

venturing units to try to create good sponsors. The presumption is that
a manager in charge of a portfolio of new businesses will have the time
to learn about them and understand the development issues they face.
In contrast, a corporate CEO or division head will often have limited
time and will be tempted to treat the new business in the same way that
he or she treats other businesses. Unfortunately, although the head of
a new businesses division may have more time to learn and more
understanding of the particular challenges involved in the development
stages, he or she usually lacks power in the broader organization (see
Chapter 8 for more discussion of corporate venturing units).
A compromise position is to have the new business report to a line
manager within the power structure and then provide extra support
from a business development function. This has been IBM’s solution in
its Emerging Businesses program. An arrangement such as this would
score yellow. It would only be possible to get a green light for the
sponsorship part of this Traffic Light if the sponsor is a line manager
with the necessary power in the organization and the appropriate skills.
In our sample of successes, 90% (where we had information)
reported to the CEO or a main division head. In just under a quarter
of these, the CEO was the initiator of the original plan. Having a
strong sponsor in the parent company appears to be as important a fac-
tor as having a strong management team at the business level.
137
Alza was a successful US west coast drug-delivery company. It

helped major pharmaceutical companies develop powerful delivery
mechanisms for the compounds they had discovered. However, like
most drug-delivery companies, Alza was ambitious to become a phar-
maceutical company in its own right. One reason was that the margins
are much bigger. Despite a number of attempts in the 1970s and 1980s,
Alza failed to make the transition. The business of acquiring com-
pounds and taking them to market just seemed too different to its core
business of adding delivery technology to other people’s compounds.
Then, in the mid-1990s, Alza hired Ernest Mario as its chairman.
Mario had been head of Glaxo Wellcome and had led Glaxo through
its fastest growth years. Within five years, Mario helped Alza do what
it had failed to do in the previous fifteen. For Mario, pharmaceuticals
was more core than drug delivery. He found it easy to sponsor and
guide the initiatives needed to help Alza make the transition. A few
years later, Alza’s success was spotted by Johnson & Johnson, who paid
a high price to acquire what Mario had created. Unfortunately, now
that Alza has achieved its ambition, the original drug-delivery business
is getting too little investment and management attention.
Another example of the power of appropriate sponsors is Rexam.
When Rolf Borjesson left PLM, a Swedish canning company, to head
up Rexam, he inherited a disparate portfolio of specialty paper and
plastic packaging businesses and an array of other activities. As he
sorted through his options he became more and more convinced that
the best plan for Rexam was to become a focused packaging company.
At an opportune moment, he bought his previous company, PLM, and
followed with other acquisitions in the canning and glass packaging
industries, becoming one of the best-performing stocks on the London
exchange during the difficult 2000–2003 period. These moves into
cans and glass were a huge leap for Rexam, but a small step for Rolf
Borjesson. In fact, he understood the canning and glass industries
rather better than the specialty packaging businesses he had inherited.
It was not difficult for him to be an effective sponsor of these new busi-
nesses in Rexam’s portfolio.
138
THE GROWTH GAMBLE
SCORING AND OVERLAPS WITH OTHER PARTS OF THE

TRAFFIC LIGHTS
The leader/sponsor Traffic Light has some overlap with the value
advantage Traffic Light. The insights of the leaders or the sponsors
may be part of the unique contribution. Also the knowledge that lead-
ers and sponsors have of the new business will affect the learning costs.
Hence there is a danger of double counting.
Experience suggests that where double counting does occur it is
more likely to be a benefit than a disadvantage. Because the people
issues are easy to underestimate when assessing value advantage, the
tendency is to give more weight to tangible factors like patents or
brands. There is, therefore, a benefit in giving extra emphasis to peo-
ple issues under the leader/sponsor Traffic Light.
An aggregate score for the leader/sponsor Traffic Light depends on
the sum of the leadership and sponsorship assessments. The leadership
element should be given greater weighting. However, a strong sponsor
can make up for some leadership weaknesses and vice versa. The
Traffic Light should not be scored as green without strong leadership
and strong sponsorship. It should be red if the leaders are clearly
weaker than the competition or if the sponsor has little influence in the
parent company or a mindset molded by a different business model. In
all other situations the assessment will be yellow.
In the eye-care and ice-cream examples the leadership/sponsorship
score would depend on the individuals involved at the time. At Mars
and Cadbury there were few, if any, managers with experience in ice
cream in Europe and none with personal insights. This would have sig-
naled a potential red light. The health and beauty company consider-
ing eye care is likely to be in a similar position unless it recruits a
quality manager from the industry or buys a high-quality smaller
competitor.
139
WHAT IS THE IMPACT ON EXISTING

BUSINESSES?
The fourth Traffic Light is about the impact the new business will have
on existing businesses. The impact might be significantly positive as a
result of customer or cost synergies (green), significantly negative as a
result of conflicts of interest and/or distraction of key managers (red),
or somewhere in between (yellow). There are two dimensions to con-
sider: synergies and distraction risks (see Box 5.7).
Box 5.7 Impact on existing businesses
Size of positive or negative synergies
Risk of distraction costs

– Operating level
– Sponsor level
– Scarce shared resources
HOW BIG ARE THE SYNERGIES?
Synergies can be positive or negative. When commercial banks entered

insurance, the benefit for the core business came from greater use of
the existing branch networks. By selling additional products and serv-
ices through these branches, the fixed cost was spread across a larger
volume of business. On the other hand, when Grand Metropolitan
owned a gaming business alongside its food brands, such as Green
Giant, there were negative synergies. The bad reputation of the gam-
ing industry undermined Grand Metropolitan’s overall reputation in
the US. The solution was to sell the gaming business.
In some situations synergies can be so important to the existing
businesses that they become the dominant logic for the new activity.
Irish Life, Ireland’s largest life insurance and pensions company,
merged with Ireland’s largest building society (savings and loans) for
140
THE GROWTH GAMBLE
the benefits it brought to the insurance business. It provided the insur-

ance business with an additional channel of distribution, the branch
network of the building society.
When Centrica, a gas and electricity utility, set up the credit card
Goldfish, the main logic was the loyalty benefits for the core business.
Competitors were introducing loyalty programs that could threaten
the core utility activities. Goldfish was Centrica’s response. It turned
out to be so successful that it quickly developed into a broader finan-
cial services business. After a few years the threat from competitors
diminished, undermining the loyalty logic for Goldfish. Without the
synergy, Centrica decided to sell Goldfish to Lloyds TSB, one of
Britain’s leading financial services companies.
When medical diagnostic imaging moved beyond classical X-ray
into CT scanners and magnetic resonance, all the major X-ray suppli-
ers like Siemens, Philips, GEC/Marconi, and GE developed into the
new business. They needed to offer magnetic resonance products to be
a credible supplier to hospital X-ray departments.
The synergy criterion is about the benefits for existing businesses.
In many cases there are no positive or negative synergies for existing
businesses. More than 75% of our database of success stories had no
significant synergy benefits for existing businesses. Moreover, it is also
comparatively rare for the synergies to be significantly negative. In
other words, this criterion is often neutral.
However, in those few situations where synergies are significant,
they can be a critical element in the Traffic Lights evaluation. In 11
cases from our database synergies were significant. In five of these,
strengthening existing businesses was the prime initial reason for
developing the new business.
In the eye-care and ice-cream examples this criterion would proba-
bly be neutral. It could be argued that there would be benefits to
existing businesses from having the brands more widely used in the
marketplace. It could also be argued that a larger company with more
businesses offers better career prospects and can, therefore, attract bet-
ter managers. However, both of these possible positive synergies are
likely to be small and can be challenged. The extra exposure of the
141
brands could dilute the image of existing businesses. The greater

spread of businesses could make the company appear less coherent to
prospective managers. Hence a neutral score is likely to be the most
easy to defend.
WHAT ARE THE DISTRACTION RISKS?
Insight 5: Managers often underestimate the loss of performance

in the core businesses that occurs when attention shifts to new
businesses and some of the most energetic managers are allo-
cated to new business projects.
Distraction risks, our way of capturing the risk of lower performance

in existing businesses, depend on two factors:
❏ The degree to which the challenges in the existing businesses

demand scarce management and financial resources.
❏ The degree to which the new business will compete for these scarce
management or financial resources.
The greater is either or both of these factors, the greater is the scope
for loss of performance in existing businesses. The distraction may not
result in an actual reduction in performance. The loss may be between
actual performance and potential.
However, we have also observed that when managers start investing
in new businesses, they often underestimate both the opportunities for
growth in their existing businesses and the threats these businesses face
from competition.
In order to understand the risks of distraction, managers need first
to understand in what areas their existing businesses most need
attention. To help managers, we have developed a checklist of ques-
tions (see Box 5.8) that identify those areas where existing businesses
most need attention. Armed with this analysis, it is easier to judge
whether a particular new business is likely to distract from the atten-
tion that the core needs.
142
THE GROWTH GAMBLE
Box 5.8 Attention tests for existing businesses
❏ Maturity test—are existing businesses fully mature (low growth, high levels of sat-
uration, low rates of innovation, and commoditized products)? If not, attention will
be needed to keep abreast of developments.
❏ Dominance test—are existing businesses dominant in their respective markets
(market share twice that of the second player)? If not, attention will be needed to
defend the existing position and win share from competitors.
❏ Profitability test—are existing businesses generating margins that are better than
competitors’, given their market and position? If not, attention will be needed to
restore profits to their rightful level.
❏ Threat test—is the horizon calm (no complaining customers, no new entrants, no
new distribution channels, no disruptive innovations)? If not, attention will be
needed to find the best response to these threats.
❏ Consolidation test—is the industry fully consolidated (three or four major players)? If
not, attention will be needed to assess merger partners and prepare for integration.
❏ Internationalization test—has the business reached its limits of internationalization
(a major player in every major market)? If not, attention will be needed to explore
different countries, find entry routes, and drive market share.
❏ Extension test—are extensions into adjacent channels, segments, products, or
value-chain positions blocked off? If not, attention will be needed to explore these
extensions.
Assessing the risk of distraction from a specific new business pro-

posal involves thinking about how the new business will compete for
operating managers, for sponsor-level managers, and for scarce
resources. At the operating level it is necessary to assess whether the
new business will be staffed by any managers from the core. If so,
which managers, how critical are they to the areas where the existing
businesses need attention, and how easy are they to replace? Often new
businesses draw on some of the most innovative and energetic man-
agers in the core, resulting in less innovation and energy focused on
the challenges in the core. The head of one business development team
explained:
143
We always end up talking about the same five or six managers. These
managers are important to the core business and are the ones most
likely to be able to make a success of our new projects.
Not surprisingly, the struggle over resources in this company was won
by the core business and, while we were doing the research, the new
businesses team was disbanded.
At McDonald’s, Partner Brands such as Chipotle and Pret A
Manger might be viewed as providing low distraction risk at the oper-
ating level. However, it depends on how these brands are managed. At
Chipotle 50% of the top team consisted of managers transferred from
the hamburger business. At Pret A Manger the percentage was lower,
but not insignificant. Moreover, both of these brands had been
acquired to see if it were possible to inject McDonald’s skills to help
them improve and grow. Hence it was intended that they would take
up the time of some managers in the core. Given the attention needed
in the core business and given the plans for brands like Chipotle, the
distraction part of the “impact on existing businesses” light would have
been red.
The second source of distraction is competition for resources at the
sponsor level in the parent company. If the person to whom the new
business is reporting has other pressing priorities in the core business
or could be usefully deployed on priorities in the core business, there
is a significant risk of distraction. In a large company like McDonald’s,
this might seem a low risk. There are large numbers of managers.
However, the top teams even at companies like McDonald’s or in divi-
sions such as McDonald’s South America are often small. If one of
these managers is spending significant time on new businesses, it will
be at the expense of investing that time in the core business.
At McDonald’s the Partner Brands were formed into a division.
This was partly to give them more attention and partly to avoid dis-
tracting too many managers. However, over five or six years, this divi-
sion had three or four different heads. Either the head was needed for
a more important job in the core business or he proved to have skills
that would be better used in another part of the business. In other
144
THE GROWTH GAMBLE
words, McDonald’s found that its Partner Brands were competing for
management talent at the sponsor level in the organization.
The third source of distraction risk involves considering how the
new business will compete for critical shared resources, like experts in
industrial design or miniaturization or software. When a high priority
is placed on succeeding in a new business, these experts are drawn away
from projects in the core business. If these skills are critical to an area
where the core needs attention, the distraction risks are high.
However, even if the skills are not part of a current priority, if they are
critical to long-term success the core can suffer imperceptibly over
time.
Between 1999 and 2005, Mercedes fell from number 3 (out of 38)
in the JD Power reliability ranking to number 28 (out of 37) and in
2004, profits from the Mercedes car marques fell 60% to levels not
seen since the late 1990s. The reason reported by the Financial Times
was “the seconding of many of its engineers to work on Chrysler’s
problems in Detroit.”6 This is a good example of the long-term impact
of a new business acquisition made seven years earlier.
We were surprised how frequently the distraction cost issue was
raised. Simon Yun-Farmbrough, then head of planning at Prudential,
first alerted us to it. He pointed out that the success of Egg had dis-
tracted managers in Prudential from making changes in the core UK
insurance business. This business was suffering from the rise of direct-
to-consumer business models. Egg was Prudential’s direct-to-con-
sumer response. However, the initial reason for launching Egg was to
help revitalize the UK insurance business. As Egg became more suc-
cessful and more independent from the UK division, it distracted
attention from the issues in the UK and began to make less direct con-
tribution to the cultural changes needed in the UK.
McDonald’s was another situation where our attention was drawn
to the issue of distraction risks. McDonald’s was so concerned about
the need to give attention to the core business at the same time as
developing new businesses that it set up two executive committees—
one for the core business and one for new businesses. This dual struc-
ture was a mechanism designed to help managers invest in new
145
businesses without distracting from the core. More recently, the issue
is still on the agenda. The Partner Brands have been placed in
McDonald’s Ventures, and Mats Lederhausen has been asked to decide
whether any of them can be significant without distracting from the
core.
At Intel distraction risks were also on the agenda. Intel defined two
priorities: Job 1, running the core business, and Job 2, investing in new
businesses. It was clear to managers that Job 1 was more important, but
work should be done on Job 2 so long as it did not risk performance at
Job 1.
What surprised us was that, despite this broad awareness of distrac-
tion costs, managers did not appear to take them into account when
assessing individual projects. The assumption was that the managers in
the new businesses should fight as hard as possible for resources.
However, although this contest would have a cost, the bigger concern
was about the impact on the new businesses rather than on the existing
businesses. Our argument is that the cost to the existing businesses can
sometimes be many multiples of the benefits from the new business,
even if it succeeds. Hence sponsors of new businesses, and even the
managers pushing individual proposals, need to make an objective
assessment of the likely distraction costs as part of the proposal to
proceed.
This objective assessment is best done by defining the main areas
where the existing businesses need attention, and then assessing the
potential for distraction at the operating level, the sponsor level, and
any levels where scarce resources are shared.
In the eye-care example, distraction risks would be likely to be low
unless the new eye-care unit draws many managers from the existing
health and beauty business or unless the health and beauty unit has
important challenges that will require the full attention of its senior
management team. In the ice-cream example, distraction risks might
be higher. Within Mars, for example, the ice-cream business would
need to draw on research, marketing, and manufacturing skills. It
would also take up significant time of a senior sponsor, probably the
CEO of Europe. Although the risks of distraction are higher in the ice-
146
THE GROWTH GAMBLE
Figure 5.9 Eye care and ice cream summaries
cream example, they are still probably in the yellow rather than the
red. With the exception of the CEO’s time, the ice-cream venture is
unlikely to be drawing on any hard-to-replace resources.
THE TRAFFIC LIGHTS IN SUMMARY
The Traffic Lights may seem a little daunting. There are four major
judgments that need to be made and each depends on a number of sub-
judgments. In our experience, however, the Traffic Lights are relatively
easy to use. Frequently it takes less than one hour to talk through the
four Traffic Lights and to arrive at a preliminary conclusion. Parts of
this preliminary conclusion may be easy to challenge, but frequently
the overall judgment—red, yellow, or green—is not in disagreement. If
it is, more work is needed and sometimes this can take days or weeks
to complete. However, it is rarely wasted work. If the judgment is
improved, a better-quality decision will result.
The two cases we have been analyzing—ice cream and eye care—
can now be summarized (see Figure 5.9). In both cases the answer is to
reject the proposal as presented. It does not make strategic sense for a
health and beauty retailer such as Boots in the UK or an upscale fash-
ion and beauty retailer such as Neiman Marcus in the US to enter the
eye-care business. Nor does it make strategic sense for Mars or
147
Cadbury to enter the ice-cream business in Europe. In both cases the

probability of earning above-average returns is too low.
In reality, Boots did enter the eye-care business in the UK using the
Boots Opticians brand. The early results were encouraging. The brand
did help generate traffic. However, in part due to being small and in
part due to learning costs, the new business found it hard to make sig-
nificant profits. Then a new competitor, with more aggressive pricing
and a franchising model, entered the business making it less attractive
for all competitors. Boots’s position as one of the weaker competitors
is now at risk. In the US, Neiman Marcus probably never even consid-
ered entering the eye-care business. As a comparison, it did enter the
gourmet food business at one point, but abandoned it after a few years
for predictable reasons.
Mars chose to enter the ice-cream business in Europe, while
Cadbury did not. With hindsight, Cadbury’s strategy of licensing its
valuable brands to an ice-cream company has proved to be more value
creating. Over more than a decade, Mars has found it difficult to make
a profit from its ice-cream sales. Long term, it may have to invest con-
siderably more effort in the ice-cream sector to gain a sustainable posi-
tion in Europe, and is unlikely to earn a return that justifies the
investment. Meanwhile, Cadbury has been earning good royalties for
the use of its brands, an earnings stream that has not required the com-
pany to invest significant additional capital.
Decision making for new businesses is often focused around a
detailed financial analysis and implementation plan. As a result, critical
evaluation of the strategic rationale can receive insufficient attention.
This is what we believe happened in the Boots and Mars examples. If
these companies had used a screening tool such as the Traffic Lights,
their strategic evaluation would have caused them to reject the project
before doing detailed financial analysis or implementation planning. It
would also have helped them develop better proposals to exploit the
advantages and resources they had.
The Traffic Lights are a set of questions that managers need to
answer that in aggregate form a strategic business case. They are
designed to be used early in the life of a project. They can be used as a
148
THE GROWTH GAMBLE
sanity check for a financial business plan, but they can also be used
before managers have developed enough information to compile a
business plan. This positioning of the Traffic Lights is important. They
can be used very early on before much information is available to assess
the potential; they can be used after some exploration and
experimentation have been done and more information is available; or
they can be used alongside a full business case analysis after detailed
research has been carried out on a new project.
The Traffic Lights do not focus on execution issues. They do not,
for example, assess whether suitable partners can be found or whether
the technology will work. These are important issues, but it is hard to
make judgments about them at the idea or even business plan stages.
The Traffic Lights presume that operational issues can be surmounted.
They focus on the soundness of the strategy rather than the ability to
execute. The execution issues are addressed in Chapter 9, when we
introduce the Confidence Check.
The Traffic Lights do not provide clear go/no go decisions for every
situation. However, they frequently give a no go answer in situations
where managers are inclined to “give it a try.” As a result, the Traffic
Lights screen out a large percentage of new business projects that
would subsequently fail, saving managers both time and money.
There should be very little in the Traffic Lights with which experi-
enced strategists or academics disagree. The intention has not been to
create some dramatic new theory. The intention has been to pull
together the fundamentals of good strategic thinking in a way that will
help managers arrive at better judgments. The insights are not major
theoretical breakthroughs. However, they are important advances in
our thinking: they were insights to us.
Despite offering no new theory, the creation of the Traffic Lights
has advanced our thinking about diversification. For readers who like
to position ideas against current theory, the next chapter reviews the
literature on diversification and positions the Traffic Lights within this
stream of thinking.
149
CHAPTER 6
DIVERSIFICATION
D
iversification has come to mean something bad. It is a word
that is associated with “overdiversification,” “conglomer-
ate,” and investments unconnected with existing busi-
nesses. When asked whether a new business project is a diversification,
managers will frequently respond in the negative: “This is not a diver-
sification, it builds on our relationships with … and it exploits our skills
at …”
Yet given the way we have defined new businesses—new business
model and separate organizational unit—all new businesses involve
some form of diversification: they all take the company into some new
markets and involve some new management skills. Hence, for those
readers who like to place ideas in context, it makes sense to look back
at the literature on diversification and position the New Businesses
Traffic Lights against other theories about what sorts of diversification
make sense.
The concept of diversification—entering new businesses—was first
discussed in the 1960s. Since then there have been two parallel tracks
of thinking about the issue. One track has focused on the vertical
reporting relationship in an organization and has explored different
views about why it would make sense for companies to have a spread of
businesses and what sort of skills the corporate-level managers need to
have in order to justify entry into new businesses. We will call this the
General Management School. The second track has focused on the lat-
eral relationships in organizations, the relationships between busi-
nesses. This track of thinking has explored different views about the
sorts of connections, overlaps, and relationships that need to exist
between businesses to justify entering a new area. We will call this the
THE GROWTH GAMBLE
Box 6.1 Schools of thought
General Management School Synergy School
General management
budgeting and planning processes Related businesses
Portfolio planning
(Boston Box) Core businesses
Value-based management Core competencies
Parenting Theory
Core, corporate-center competencies
Synergy School. Frequently the different tracks become intertwined,

but it is worth distinguishing between them for purposes of greater
understanding (Box 6.1).
GENERAL MANAGEMENT SCHOOL
In the 1960s many companies, such as ITT in the US and Slater

Walker in the UK, diversified widely. The logic at the time was “gen-
eral management skills.” These companies believed that they had
developed an understanding of how to make money in business, and
skills at planning, budgeting, performance monitoring, and control
that enabled them to buy almost any underperforming business and
tune it up.
151
DIVERSIFICATION
Their thinking was given credence by the rise of the business

schools and the belief that management was a profession that could be
learned. Good managers had, it was thought, the skills to manage any
business. Initially, therefore, the logic for diversification was one of
performance improvement through good management. Managers in
the parent company knew how to improve the performance of an
acquired business or how to help set up a greenfields business so that
it would succeed.
This thinking became hard to refute as the conglomerates of the
time, such as ITT, outperformed most other companies. As a result,
many companies followed the same logic and started diversifying.
The recession of the 1970s brought this diversification trend to an
abrupt stop. Companies ran out of money, and were faced with the task
of deciding which businesses to keep and which to close or sell.
Helping them was a consulting company called Boston Consulting
Group (BCG).
Building on work BCG had done on why companies with leading
market shares make the highest profits, its consultants developed the
now famous Boston Box (Figure 6.1). This, the first of the consultant’s
matrices, asked companies to plot their portfolio of businesses against
two axes: their relative market share and the market’s growth rate.
Based on these two inputs, the matrix categorized each business as cash
cow, dog, question mark, or star. Companies were advised to keep their
cash cows, because these provided the funds to fuel growth, keep the
stars, because these provided today’s growth opportunities, and invest
in a few question marks and new businesses with star potential, which
would provide tomorrow’s growth. Dogs and doubtful question marks
could be sold or closed.
The logic for being in multiple businesses was still based on general
management skills: any business was acceptable. What the Boston Box
added was a financial logic and a growth logic. Companies can provide
a service to shareholders by putting together a portfolio of different
businesses that produce a balanced, stable performance record. When
choosing which businesses to include in the portfolio, companies
should try to maintain a balance: some cash cows to fund expansion,
152
THE GROWTH GAMBLE
Figure 6.1 The growth share matrix
some stars to deliver today’s performance, and some investments in

“businesses of the future” to deliver tomorrow’s performance. The
growth challenge for managers was thought to be about identifying
businesses of the future and maintaining a reasonable portfolio
balance.
Excited by the rigorous analysis supporting the Boston Box, many
companies applied it not only to cull their existing portfolio but also to
define target companies to acquire or target markets to enter. Companies
with high growth prospects looked for cash cows to help fund their
growth. Companies with low growth prospects looked for high-growth
businesses that could be propelled into a leadership position.
By the mid-1980s, however, disillusionment had started setting in.
Researchers were noting that the best performers in most industries
were companies like Coca-Cola, McDonald’s, or Boeing, which were
not diversified: companies that stuck to one business.1 At the same
time, managers in diversified companies were finding it hard to man-
age their wide spread of businesses. Diversified companies were
underperforming focused companies. Some were even choosing to
spin off businesses and break up their portfolios. In addition, the
finance theorists were arguing that shareholders could spread risk
153
DIVERSIFICATION
more easily than companies. It made no sense, they argued, for com-
panies to diversify in order to produce a portfolio with a stable per-
formance. Shareholders could do it more cost effectively themselves.
By the early 1990s, with the collapse of the last committed con-
glomerates such as Hanson, the General Management School of think-
ing had become completely discredited and the Synergy School had
become dominant. Nevertheless, with tools like value-based manage-
ment, the idea that corporate-level managers could develop skills that
are appropriate to a wide range of businesses was still alive.
THE SYNERGY SCHOOL
Like the General Management School, the Synergy School has deep
roots. The idea was that some businesses are sufficiently similar or
have sufficient connections that they can be or need to be managed by
one company. The concept was discussed in the 1960s by Igor Ansoff,
one of the earliest business strategists.2 He invented the product/mar-
ket matrix, suggesting that managers should avoid diversifying into
businesses that involve both new products and new markets (see Figure
6.2). He also invented the term synergy as applied to business.
In the early 1970s, Richard Rumelt published a study of diversifica-
tion giving factual support to Ansoff’s ideas.3 He noted that over 50%
of companies had diversified into multiple businesses. However, when
he correlated performance against type of diversification, he found that
the “related diversifiers” performed best. The implication was that
companies should diversify into businesses that are related in product,
market, or skill terms. The idea of relatedness caught on. Those com-
panies that were not diversifying based on a belief in general manage-
ment skills were actively looking for related businesses.
The process of defining related businesses begged the question
“related to what?” The answer was “related to our core businesses.”
But this required companies to define their core businesses. Hence this
school sparked a process of defining core businesses in order to be able
to identify related businesses.
154
THE GROWTH GAMBLE
Figure 6.2 The product/market matrix
The development of the resource-based view of strategy in the

1980s gave a further boost to the Synergy School. This argued that
success in any industry goes to the company that has the most appro-
priate resources in terms of location, relationships, assets, technology,
skills, and other factors. Since commonly available resources, such as
skills that can be hired easily from the marketplace, do not distinguish
one company from another, it is the unique resources a company pos-
sesses that drive its success or failure.
When the resource-based view was applied to the issue of diversifica-
tion, Gary Hamel and CK Prahalad came up with the idea of core com-
petencies.4 Successful diversified companies, they argued, are companies
that possess an operating competence important to the success of all of
their businesses. The competence might be a certain kind of marketing,
such as P&G’s skill at mass-market consumer goods. It might be a tech-
nology, such as 3M’s knowledge of thin film coatings. Any competence
that is important to the success of a number of different businesses could,
Hamel reasoned, provide the logic for diversification. To decide which
new businesses to enter, managers should identify their core competen-
cies and spot businesses where these competencies add value.
155
DIVERSIFICATION
The popularity of the core competencies idea was in part due to the
fact that it provided a potential answer both to the question of what is
core and to the question of what is related. It seemed compelling at the
time.
PARENTING THEORY
Work by Michael Goold, Andrew Campbell, and Marcus Alexander in

the early 1990s attempted to bring the two schools together into one
theory of diversification.5 Parenting Theory stated that general man-
agement skills, such as planning and control, could not provide the
logic for diversification decisions because, by definition, the skills are
generally available. This undermined part of the logic for the General
Management School.
However, special and not generally available skills do exist at the cor-
porate levels of some companies. For example, financing mines in awk-
ward places at awkward times in the cycle and with awkward political
risks is a skill possessed by the mining company Rio Tinto. This helped
it diversify into a range of mining industries. The skill of orchestrating
a cadre of product managers across an international network helped
Unilever diversify into more and more countries and expand its port-
folio of fast-moving consumer products. Companies with unique corpo-
rate-level skills and other resources, such as brands or relationships or
patents, can, Goold and colleagues suggested, diversify into new busi-
nesses where these skills or resources will add value. Parenting Theory
acknowledges the importance of corporate-level skills. However, rather
than focusing on the general management skills at the corporate center,
Parenting Theory focuses on the unique skills that one corporate center
has compared to others. It is these unique skills that justify one parent
owning a business rather than another.
Parenting Theory also challenged the basic ideas in the Synergy
School. Parenting Theory stated that lateral relatedness between busi-
nesses was not in itself enough to justify diversification. There were
two challenges to the relatedness thesis. First, synergy potential is not
156
THE GROWTH GAMBLE
a sufficient reason on its own. Synergy, in the form of alliances and

coordinated working, can be achieved between independent compa-
nies. Coca-Cola can have a joint marketing campaign with
McDonald’s without owning McDonald’s. Hence relatedness only
offers a logic for diversification if the synergy opportunity requires
joint ownership.
If the synergy benefits can be achieved without joint ownership, an
alliance or licensing deal or trading relationship is all that is needed.
Joint ownership is only required for synergies that cannot be achieved
through arm’s-length arrangements. For example, if one company
believes that it can improve the performance of another company by
imposing its strategy on the other, joint ownership is normally needed.
Also, if one company believes that there is potential to cross-sell prod-
ucts with another company, but the other company does not, joint
ownership is required to release the potential. Only some kinds of syn-
ergy require joint ownership. Therefore, only some kinds of synergy
justify diversification.
The second challenge to the Synergy School was that, even when
there is a synergy logic that requires joint ownership, it may not be suf-
ficient to justify joint ownership. For example, in the aluminum indus-
try, most companies in smelting have tried to get into fabrication and
other downstream businesses. Not only are these businesses less capi-
tal intensive, offering higher returns on assets, but they can pull vol-
ume through the fixed-asset smelting business.
In practice, almost all the attempts failed. The reason for the failure
was not the lack of synergies from pulling through extra volumes, but
the different management skills required in the downstream busi-
nesses. The upstream parent proved to be “incompetent” at managing
the downstream businesses, which lost money or underperformed. The
implication is that lateral synergies can only justify diversification if
ownership is necessary and if the management approach needed in the
new business is one that the parent managers understand. This nor-
mally means that the management approach that is best for the new
business needs to be similar to the approach applied in the existing
businesses.
157
DIVERSIFICATION
Putting together these thoughts about top-down skills and lateral

synergies, Parenting Theory defines three conditions that need to exist
to justify entering a new business.
❏ First, the business must be underexploiting its potential in some

way: there must be an opportunity to improve it either through top-
down management guidance or through synergies.
❏ Second, the parent company must have the appropriate top-down
skills or synergy capability so that the improvements can be
achieved.
❏ Third, parent-company managers must understand the manage-
ment approach needed in the new business well enough to be able
to adjust normal corporate policies and rules of thumb so that the
new business is not inadvertently damaged.
Parenting Theory focuses on the role of the parent managers. It is

these managers who will deliver the top-down guidance, orchestrate
the necessary synergies, or damage the new business with inappropri-
ate guidance, people decisions, or policies. The alignment between the
General Management School and the Synergy School is achieved by
recognizing that there are top-down reasons and synergistic reasons
for entering new businesses. Parent managers play a vital role in both.
Whereas the Synergy School focuses on core operating competencies
and the General Management School focuses on corporate-level skills,
Parenting Theory combines the two in focusing on the unique skills of
parent managers: the core, corporate center competencies (see Box
6.1).
Like both schools, Parenting Theory is a competitive theory.
Companies should be aiming for a “parenting advantage”: their core
competencies at the corporate level should be superior to those of rival
parent companies.
158
THE GROWTH GAMBLE
ASHRIDGE PORTFOLIO DISPLAY
Parenting Theory is summarised in the Ashridge Portfolio Display

(Figure 6.3).6 The axes of the display capture both the potential for
parent managers to add value to a business and the risk that they will
inadvertently harm the business. Top-down guidance or synergy man-
agement is part of the value-added axis. This is where Parenting
Theory has brought together the ideas from both the General
Management School and the Synergy School. The other axis, the risk
of value destruction, is largely overlooked by both schools. One of the
contributions of Parenting Theory is to give attention to this value-
destruction axis.
Businesses that fall in the top right of the matrix are ones where the
parent company can add a lot of value and there is little risk of value
destruction. For Unilever, these would be fast-moving consumer goods
businesses in new geographies but in product categories where
Unilever already has other businesses. For Exxon, these would be oil
exploration, refining, or distribution businesses in geographies where
Exxon can link to existing assets. These “heartland” businesses are the
company’s core businesses.
“Edge-of-heartland” businesses are ones where either there is less
opportunity for the company to contribute value or more risk of value
destruction. They are businesses where the parent managers will need
to learn some new skills in order to succeed. For example, when Mars
entered Russia after the collapse of the Communist bloc, it was enter-
ing a marketplace of which it had little previous knowledge. The edge-
of-heartland section of the matrix is the equivalent to the new
product/same market or new market/same product sections of Ansoff’s
product/market matrix.
“Value traps” are businesses where there is an opportunity to add
value, yet the management approach needed to make the business a
success is so different from the experience of parent managers that
there is a big risk of value destruction. Aluminum fabrication busi-
nesses, for example, were value traps for aluminum smelting
companies.
159
DIVERSIFICATION
Figure 6.3 The Ashridge Portfolio Display
When Unilever acquired Elizabeth Arden, an up-market cosmetics

company, it was venturing into a value trap. While the product
development, manufacturing, distribution, and research activities of
Elizabeth Arden were all familiar to Unilever, the pricing and market-
ing of the products to up-market customers was less familiar. In the
early years of ownership, Elizabeth Arden’s performance improved sig-
nificantly as Unilever’s skills in areas such as research and product
development influenced Elizabeth Arden managers. However, over a
decade, the mismatch between the management approach for a mass-
market company like Unilever and that needed for an up-market com-
pany like Elizabeth Arden caused problems.
After more than 10 years of ownership, Unilever sold Elizabeth
Arden in a private equity deal. Within a couple of years, the managers
of Elizabeth Arden had changed many of the policies from the previ-
ous era, significantly accelerated growth, and improved profits.
Despite being able to add value to Elizabeth Arden in the early years,
160
THE GROWTH GAMBLE
Unilever’s investment was not a success. The skills of managing up-

market brands were sufficiently alien to most Unilever managers that
they inadvertently gave some wrong guidance and imposed some
wrong policies.
“Alien territory” in the Ashridge Portfolio Display includes busi-
nesses where the company has nothing to contribute and the manage-
ment skills needed are unfamiliar. For Unilever, aluminum smelting
would be in alien territory, as would fast-moving consumer goods for
an aluminum company.
“Ballast” refers to those businesses parent managers know well, but
are able to add little to. Normally these are businesses that have been in
the company for a long time. The managers at the parent level probably
spent most of their early years in these businesses and ran them before
rising to the corporate level. The corporate guidance and policies are
aligned to these businesses, but the businesses are already well managed:
there is little opportunity to add value. The margarine businesses in
Unilever might be ballast, as might US oil exploration for Exxon.
The Ashridge Portfolio Display gives guidance about the sorts of
businesses a company should consider entering. First, it should enter
businesses in its heartland: businesses where it has a lot to add and
where it understands the management challenges. Second, if there are
no heartland opportunities, the company should consider edge-of-
heartland opportunities: businesses where the ability to add value is less
certain or the knowledge of the challenges facing managers less com-
plete. Third, the company should avoid businesses where it can con-
tribute little or where its understanding of the management challenges
is weak.
Does the Ashridge Portfolio Display, published in 1994, say more
than Ansoff’s product/market matrix, published in 1965? Yes and no.
Both matrices suggest that managers should stick to businesses they
understand and diversify into new businesses cautiously. Ideally they
should build on knowledge they already have and avoid businesses
where everything needs to be learnt afresh. The difference is that the
Ashridge Portfolio Display emphasizes different dimensions of know-
ledge. Parenting Theory focuses on the skills of parent managers and
161
DIVERSIFICATION
on the management approach that these managers find most natural,

while the Ansoff matrix focuses on knowledge about the markets or the
products.
Clearly, the products and markets that form the current core busi-
nesses will have influenced the skills of the parent managers and the
management approach that they find most natural. In other words,
there is a link between Ansoff’s thinking and the Ashridge thinking. But
the focus of the latter is the management approach that the parent man-
agers find most natural and how appropriate it is for the new business.
Some products or markets may seem similar using the Ansoff matrix,
for example aluminum fabrication for an aluminum smelting company
or up-market perfume for Unilever, yet require a rather different man-
agement approach and involve some different management skills.
The Ashridge Portfolio Display also focuses on added value. Given
that there will be some learning costs in any new business, it is impor-
tant that the parent company brings some contribution to the new
area. Without some value added, the parent company will find it hard
to earn back its start-up costs, acquisition premium, or learning costs.
Ansoff’s matrix presumes that management will be able to add value to
businesses in similar product or market areas. This is not necessarily
the case.
The final point to make with regard to the Ashridge Portfolio
Display is the link between the display and the research project that led
to this book. The research was aimed at understanding how some com-
panies successfully entered businesses that are not in their heartland or
edge of heartland. How did Mannesman, a German engineering com-
pany, become a major mobile phone operator? How did Dixons, an
electrical goods retailer, become the leading internet service provider
in the UK? How did GE, a manufacturing company, become one of
the strongest global players in financial services? How did DSM, a
mining company, transform itself into a competitor in the life sciences
industry? As a result, our database of successes is skewed to situations
in the edge of heartland or outside. Most of the new thinking con-
tained in this book is therefore about how companies diversify into
areas that are not obvious.
162
THE GROWTH GAMBLE
LINKING PARENTING THEORY AND THE TRAFFIC

LIGHTS
The Traffic Lights, we believe, have taken the thinking about diversi-
fication decisions a step further forward. They build on Parenting
Theory but, informed by our research, they add some important new
elements and give additional focus on some others (see Box 6.2).
First, Parenting Theory has a lot of links with the value advantage
section of the Traffic Lights. The “valuable contribution” element is
similar to Parenting Theory’s added-value concept. What does the
company bring to this new business? However, the thought in the
Traffic Lights is more inclusive than the added-value concept in
Parenting Theory. Whereas Parenting Theory gives special attention
to the contribution of the parent levels, the Traffic Lights just demand
that there is some significant contribution. For example, when Boots
entered the wellbeing services business, one of the contributions was
spare space at some of its drugstore sites. This kind of contribution is
not easy to label as parenting added value. By using the looser term
“valuable contribution” and recognizing that this contribution can
come from operating levels or parenting levels, the Traffic Lights have
an advantage over Parenting Theory.
“Tradability” is an element particular to the Traffic Lights. This is
a significant addition to previous thinking. Since the reason for enter-
ing a new business is to create additional value, the Traffic Lights
point out that a new business is only needed if the value cannot be cre-
ated through a less risky approach: a sale, a license, an alliance, or a
joint venture. Parenting Theory does not include an equivalent
concept.
When taken together, the “contribution” and “tradability” variables
often arrive at an answer that is similar to the “added value of the par-
ent” variable. This is because many types of contribution, such as the
spare space in Boots’s drugstores, can be traded for value. In other
words, a contribution is frequently discounted because it can be turned
into value without entering a new business. Contributions that are hard
to trade are often skills or knowledge, particularly corporate-level skills
163
DIVERSIFICATION
Box 6.2 Traffic Lights and Parenting Theory
Elements of Traffic Lights Parenting Theory coverage
❏ Value advantage ❏ Addressed directly, but

– Unique contribution – focused on parent level
– Tradability – not included
– Contribution of rivals – included
– Learning costs – focused on parent level
❏ Profit pool potential ❏ Not included
❏ Leaders/sponsors ❏ Focus is on sponsor more than leader
❏ Impact on existing businesses ❏ Does not include distraction risks
and knowledge. As a result, sources of parent added value are often

among the strongest reasons for entering a new business.
The “contribution of competitors” is taken into account in all
theories that acknowledge the competitive nature of business.
Parenting Theory, like the Traffic Lights, explicitly recognizes the
importance of rivals: the value added by the parent needs to be greater
than that of competitors.
The “cost of learning the new business” is another element where
the Traffic Lights offer a more complete picture than Parenting
Theory. The Traffic Lights include the cost of the learning that will
need to be done by the managers leading the new business unit, as well
as the cost of the learning that will be required by parent managers.
Parenting Theory only explicitly addresses the latter. Parenting
Theory presumes that business-level managers will know their busi-
ness well enough. If the new business is an acquisition, this may be the
case. However, if the new business is a greenfields operation, the man-
agers in charge may have little experience in the new area.
Second, the Traffic Lights include a section on the profit pool
potential. This section is not part of Parenting Theory, in the same
164
THE GROWTH GAMBLE
way that it is not part of Ansoff’s matrix or core competence thinking.

These theories are all focused on the fit issue: does this new business
fit with the company? Clearly, however, fit is not the only issue. As
every manager knows, there is also the question of how attractive the
new market is. In fact, for most managers this is a more important
question than the fit one.
As we explained in Chapter 5, most markets offer the opportunity
to earn a good return, at least for good competitors. However, some
markets do not, and these “dogs” need to be avoided. Also, some mar-
kets are so attractive that they offer good returns even for weaker play-
ers. These “rare games” may provide the opportunity for a
diversification that would not be considered for other reasons.
Third, the leadership/sponsorship section is unique to the Traffic
Lights, but it is connected both to the concept of value destruction that
is central to Parenting Theory and to the concept of competition.
Unless the leaders and sponsors of the new unit are strong individ-
uals who are trusted by the rest of the organization, the new unit is
liable to be subjected to the rules and policies applied to the other busi-
nesses. Hence value destruction by parent managers, a frequent danger
when developing a new business model, is best kept in check by having
strong leaders and sponsors who can fight off the well-meant, but
deadly, influences.
Strong leaders and sponsors are also essential for winning in the mar-
ketplace. Without them, a new business has little chance of taking advan-
tage of changes in the market or weaknesses in competitors. Strong
leaders and sponsors could be thought of as part of the contribution or
added value that the company brings to a new business. However, lead-
ers and sponsors are treated as a separate term in the Traffic Lights
because the research exposed how critical they are to success.
Fourth, “impact on existing businesses” is another term that is
partly unique to the Traffic Lights. The Parenting Theory concept of
added value includes positive and negative synergies, but it does not
include distraction risks. The research that led to the Traffic Lights has
highlighted this element, suggesting that special attention should be
given to the impact of distraction risks.
165
DIVERSIFICATION
In summary, the Traffic Lights include some ideas that are not part
of Parenting Theory, such as tradability, profit pool potential, and dis-
traction risks. They also give some added focus to issues such as unit
leadership and operating-level contributions. As a result, the Traffic
Lights give more clarity to the edge-of-heartland section within
Parenting Theory and point at two types of diversification that would
be overlooked by Parenting Theory: saplings and rare games.
SAPLINGS AND RARE GAMES
In our database of 50 successes, a number of the successes could not be

classified as either heartland or edge of heartland. In fact, the database
was put together in order to understand these “outside-heartland”
successes.
Saplings are a type of new business opportunity. Saplings are activ-
ities that are already part of the company, but are not heartland or
edge-of-heartland activities. In fact, they are often not thought of as
separate businesses. They exist for two prime reasons. They were cre-
ated in support of some core business or they were acquired along with
some core activity and never sold. For example, Hewlett-Packard
started making its own computers in the 1960s because it was finding
it hard to get a reliable supply of minicomputers to support systems
using its test equipment. In the 1970s, the unit started selling mini-
computers for use with other technical systems not involving its instru-
ments. The management of the unit frequently tried to persuade the
parent company to allow it to enter the broader computer business for
general-purpose commercial applications. However, HP was reluctant
to take on IBM and other large mainframe companies. Then in the
early 1980s, as IBM was dismantling its research into RISC technol-
ogy, HP decided to acquire IBM’s research team and enter the general-
purpose computer business. Over the following 20 years, the computer
business grew 25-fold and became the dominant part of HP.
Another example is from the drug industry. Boots, a British retailer
of drugs, also owned a pharmaceutical business. This pharmaceutical
166
THE GROWTH GAMBLE
business primarily produced prescription drugs, but it also manufac-

tured a few over-the-counter products for sore throats and minor ail-
ments. When Boots sold its pharmaceutical business, it retained these
over-the-counter products and decided to set them up as a separate,
globally focused consumer products business. The products had never
been given much marketing support by their pharmaceutical parent
and Boots was able to build a substantial business in over-the-counter
consumer products.
These are two examples of how an activity that exists for one reason
can sometimes be turned into a significant new business, even though
it is not in heartland or edge-of-heartland territory. Other examples
include GE’s financial services business and GrandMet’s drinks
business.
Just being an existing activity is not enough. The special quality of
a sapling is that it has a very strong management team. Often the man-
agement team of this activity has been trying to raise the status of the
activity within the portfolio, but has been turned down or ignored.
This was true in both Hewlett-Packard and Boots. The sapling anal-
ogy is appropriate here. Think of a young tree that has been cut back
two or three times but is still growing vigorously.
The Traffic Lights explain why saplings make sense. Saplings will
normally get a yellow light for value advantage even though they are not
heartland or edge-of-heartland businesses. This is primarily because
they will have low learning costs. Because the activity has been part of
the portfolio for a few years, the learning has already happened.
Saplings will get a yellow light for profit pool potential, as do most busi-
nesses. Saplings get a green light for leadership/sponsorship because
they happen to have unusually strong management teams. Finally,
saplings get a yellow or green light for impact on existing businesses
because they normally exist originally to support the current businesses.
Based on our database, the sapling category is important. It
accounts for about 15% of our sample (see Figure 6.4 overleaf). The
implication is that companies seeking new businesses should look
around their existing activities to see if any of these are potential
saplings. They do not have to look very hard. Because saplings have
167
DIVERSIFICATION
Figure 6.4 Parenting logic
strong and pushy management teams, they are likely to be noisily com-
municating their ambitions already.
The problem with saplings is that, as they grow, they will not sit
easily with the existing businesses. If they do, they are more edge of
heartland than sapling. More normally, they either take over the nest,
cuckoo-style, or they are subsequently sold. In the case of Hewlett-
Packard, computers took over the nest and turned the company into a
very different organization both in portfolio and cultural terms. At
Boots, the drug-retailing business is so dominant that the consumer
products business, Boots Healthcare International, will almost cer-
tainly be sold at some point. At GrandMet the drinks business became
the dominant division and provided the logic for the merger with
Guinness. Subsequently all the other businesses have been sold. At GE
financial services has become the biggest generator of profits and is
likely to be demerged at some point.
Rare games are the second type of new business opportunity that
falls outside heartland and edge of heartland. Rare games are those few
business opportunities that will reward almost anyone who invests at
168
THE GROWTH GAMBLE
the right time: bulk chemicals in the early 1970s, personal computers
in the early 1980s, telecommunications in the early 1990s.
Rare games exist when demand outstrips supply for a significant
number of years and when the main advantage comes from investing
early. Companies who invest early have written down assets by the time
others enter the market. This enables them to earn a decent return on
their investment, even when supply catches up with demand.
Rare games get a yellow light for value advantage because the advan-
tage of early investment balances the disadvantages of learning costs.
Moreover, they are normally “new-to-the-world” opportunities, such as
internet service provision, where no company has any obvious value
advantage. They get a green light for profit pool potential. They get a
yellow light for leadership/sponsorship, so long as the company can find
quality managers to lead the effort. Finally, they get a reddish yellow light
for impact on existing businesses, because they are normally unrelated to
existing businesses and require different managers.
In our sample, around 15% of the successes are rare games at least
in part. Hence, even though rare games account for probably less than
5% of the opportunities, they are an important category for a company
looking to diversify.
DISRUPTIVE TECHNOLOGIES
A category of new businesses that has been much discussed in the last
five years is the so-called disruptive technologies. This term was coined
by Clayton Christensen and means that either the product technology
or the business model is difficult for existing competitors to copy.
Christensen argues that many of the most significant new businesses
are successful because they benefit from disruptive technologies.
In this section we will compare the Traffic Lights criteria with
Christensen’s disruptive technologies criteria to show that the Traffic
Lights are more comprehensive. Those readers unfamiliar with
Christensen’s work or uninvolved in technology-led industries may
want to skip to the end of the chapter.
169
DIVERSIFICATION
Christensen identified two types of disruptive innovations:
❏ Low-end disruptions that start with a technology better able to

serve the bottom end of the market. Over time as this new technol-
ogy develops, it gradually becomes competitive in more and more
up-market segments.
❏ New-market disruptions that start with a technology or business
model that can serve the “non-consumers,” potential customers
who are excluded from the market because current products do not
meet their needs or are too expensive. As the technology develops,
it also becomes more and more competitive and gradually wins
share from the mainstream.
In his latest book, Christensen argues that disruptive innovations are

the best way to develop significant new businesses. Companies looking
to create a significant new business should focus on looking for and
then organizing to support disruptive innovations.
Christensen suggests that low-end disruptive opportunities arise
because companies pursue product improvement and innovation, often
outstripping the needs and desires of many customers. Customers at
the low end thus become overserved, and are often unprofitable as a
result. Innovation at the low end, providing simpler and cheaper prod-
ucts, can be very attractive to newcomers. If the innovation is suffi-
ciently low cost it can be highly profitable. Incumbents have no
immediate incentive to retaliate since price retaliation could under-
mine more profitable segments and their least profitable customers are
being targeted; indeed, losing these customers is often beneficial in the
short term. Thus margins remain high for the traditional incumbents.
The real threat to incumbents is that the technology used by the
newcomer improves. Over time it becomes competitive in more and
more segments. Steel mini-mills are a classic example, entering in the
late 1960s via reinforcing bars (poor quality, low cost, highly unprof-
itable products for major integrated steelmakers), only to innovate
upward through the product range, product by product, until they now
dominate the industry.
170
THE GROWTH GAMBLE
In addition to new-market disruptions, Christensen also distin-

guishes between sustaining innovations and disruptive innovations.
Sustaining innovations are those that can be incorporated as part of a
company’s existing business model. Incumbents nearly always succeed
against newcomers if the innovation is sustaining and nearly always fail
if the innovation is disruptive. Hence, companies wanting to enter new
businesses should look for disruptive innovations.
In summary, he identifies three types of innovation:
❏ Sustaining innovation, which can be incremental or breakthrough

performance improvement in attributes valued by the industry’s
most demanding customers, targeted at the most profitable cus-
tomers most likely to pay for improvement, and improves profit
margins by exploiting existing processes and cost structures, making
better use of current advantage.
❏ Low-end disruption, which provides good enough performance at a
lower price, targeted at lower-end, overserved, often unprofitable
segments with a new operating or financial business model that is
profitable on low price.
❏ New-market disruption, which provides lower performance in tradi-
tional attributes but higher in new attributes (e.g., simplicity and
convenience), is targeted at non-consumption (typically customers
without the money or skills), and with a business model that makes
money at lower unit prices and volumes.
Christensen’s advice to companies to look for disruptive innovation

opportunities is good counsel. These opportunities can produce remark-
able performance as the disruptive innovation gradually wins out against
the incumbent players. However, disruptive innovation opportunities
account for less than 10% of our success stories. Christensen would argue
that this is because managers have only recently understood the enormous
potential of focusing on disruptive opportunities. In the future, he would
predict that a higher percentage of successful new businesses will come
from following a disruptive innovation strategy. As a result, we thought it
would be useful to compare his advice with our Traffic Lights concept.
171
DIVERSIFICATION
CHRISTENSEN’S CRITERIA
Christensen poses a number of tests to help identify disruptive oppor-

tunities. For a new-market disruption:
❏ Is there a large population who have not had the money, equipment or skill
to do this for themselves, so have gone without or paid someone else to do
it?
❏ To use the product, do customers need to go to an inconvenient central
location (i.e. involve high search costs)?
For a low-end disruption:
❏ Are there customers at the low end who would be happy to purchase a less
good (good enough) performance at a lower price?
❏ Can we find a profitable business model at discount prices?
In addition, to help managers decide whether an innovation is sustain-

ing or disruptive:
❏ Is the innovation sustaining to at least one significant incumbent, so giv-

ing them an immediate incentive to retaliate?
DISRUPTIVE TECHNOLOGIES AND VALUE ADVANTAGE
Christensen points out that any company in possession of a disruptive

technology has a huge advantage compared to the incumbents using
the old technology: the unique contribution is likely to be large relative
to the other terms in the Value Advantage Equation. However, he
overlooks most of the negative items in the equation. He does not dis-
tinguish between innovations, such as patents that can be “traded,” and
those, such as new business model ideas, that require the owner to
enter the new business. Also, he does not explicitly recognize the dan-
gers of other companies entering the business with the same technol-
ogy, but with other advantages to bring to the market. Finally, he does
172
THE GROWTH GAMBLE
not address the learning cost issue. One of the reasons disruptive inno-
vations cause incumbents to lose out to new entrants is the learning
costs. The incumbents need to unlearn the rules that apply to their
current business before they can fully learn the rules that apply to the
new business.
DISRUPTIVE TECHNOLOGIES AND THE PROFIT POOL
Christensen’s criteria are not easily compared with our Traffic Lights.
On the issue of profit pool potential, disruptive innovations are likely
to have attractive business models, although this may be less evident at
the early stages of the new technology. As the technology improves, the
business model becomes more and more attractive. Hence when deal-
ing with disruptive technologies, it is important to look ahead at the
business model that may be possible once the technology has devel-
oped. Disruptive technologies are also likely to be high-growth oppor-
tunities and to present the potential for leadership.
In addition to a focus on disruptive innovations, Christensen also
argues for managers to be “impatient for profit” and “patient for
growth.” This suggests that he, like us, favors opportunities where the
cost of proving the potential is small relative to the ultimate market.
Christensen does not address directly the five forces criterion. In
our experience, this is a common oversight. Managers are drawn by the
size of the opportunity and forget that promising opportunities often
attract many competitors in a race that makes it hard for any to earn a
decent return. The telecommunications business in both mobile
phones and fiberoptics is an example.
Christensen also does not explicitly focus on vulnerabilities. Some new
technologies are particularly vulnerable to developments by partners or
complementers. Also, improvements in existing technologies often occur
in a way that undermines the profitability of the new technology.
Christensen’s disruptive technologies are therefore likely to score
well on profit pool potential unless they involve technologies that are
made available to many competitors or are particularly vulnerable. In
fact, if it is possible for a company without the new technology to enter
173
DIVERSIFICATION
the new area, it is likely to be unattractive from a five forces perspec-

tive. This suggests that most disruptive opportunities would score
either green, because a few competitors have hard-to-copy technolo-
gies, or red, because the technology is available and likely to attract
many competitors.
DISRUPTIVE TECHNOLOGIES AND LEADERS/SPONSORS
Christensen does not given enough attention to the role of leaders or

sponsors of the new business. His focus is on the disruptive innovation
and he presumes that appropriate leadership will accompany the inno-
vation. Often this is not the case. There are many examples of an inno-
vation being developed in a particular company or country only to be
exploited by a different company or country. This is because the inno-
vator did not have the leadership skills to develop the business.
Christensen gives more attention to the sponsor issue. He recog-
nizes the danger of inappropriate sponsorship and recommends that
disruptive innovations should be set up in separate divisions reporting
at the same level as the businesses with which they are competing.
Again, however, he presumes that the manager to whom the new busi-
ness reports can be an appropriate sponsor. While he provides ample
evidence of parent companies whose influence on a new disruptive
technology is negative, he presumes that the solution is for them to
understand the importance of disruptive technologies. Our experience
suggests that the causes of negative parenting are often much deeper
and harder to overcome than Christensen suggests. Given that these
managers will almost certainly have grown up in the rival business, it is
unlikely that they will be good sponsors.
DISRUPTIVE TECHNOLOGIES AND THE IMPACT ON THE

CORE
Christensen acknowledges the negative synergies implied by disruptive

technologies, pointing out that these are often one of the reasons for
incumbents failing. He correctly argues that incumbents should ignore
174
THE GROWTH GAMBLE
these negative synergies because they will happen anyway. If the

incumbent does not cannibalize its own business, some competitor will
gain market share.
Christensen, however, overlooks the distraction risks. By launching
and supporting a rival business model, the company may be hastening
the collapse of its existing businesses in a way that would not happen if
it focused on the incumbent business model. Certainly a paranoid pur-
suit of every new innovation, in case it proves to be a disruptive tech-
nology, is likely to lead to underperformance when compared to a
company that sticks to its core business.
In analyzing our case studies of success, only three seem to fit the
disruptive model: IBM with its PC, Dixons and Freeserve, and the
Royal Bank of Scotland and Direct Line. The rest relate to a variety of
other insights, for example a sustaining innovation, the acquisition of
assets at an attractive price, or opportunities from discontinuous
change such as new market regulation or other political events.
Although car insurance business Direct Line clearly has been a
highly successful disruptive business model, it does not entirely fit the
Christensen template. It was not targeted at the low end of the market.
It exploited better information about risk and the realization that the
least risky, and so most profitable, customers were cross-subsidizing
poorer risks. Wood and his colleagues at Direct Line thus focused only
on this segment, using a low-cost operational model. Incumbents had
every reason to retaliate, and eventually did so. Direct Line’s phenom-
enal profitability was badly damaged for a period, but it had created a
sufficiently strong position to ride the storm and come through to be
the UK’s biggest and most profitable car insurer.
As well as the three examples that broadly fit the disruptive inno-
vation model, three have been the direct consequence of disruptive
innovations in adjacent markets. UK retailer Dixons has ridden the
PC disruption with PC World, and the mobile telephony disruption
with The Link. However, the new retailing concepts were sustaining
innovations for Dixons. Acer, a Taiwanese computer manufacturer,
made the transition from games modules to major computer company
through early moves in PC clone manufacture. Although the PC was
175
DIVERSIFICATION
a disruptive innovation for IBM, it was a sustaining innovation for

Acer.
In conclusion, Christensen’s concept of disruptive innovations is
useful and suggests some additional questions that managers should ask
when assessing the attractiveness of the business model and the size of
the unique contribution. However, identifying a disruptive innovation
is neither necessary nor sufficient for those seeking significant new
businesses. Once a disruptive opportunity has been identified, man-
agers still need to put it through the Traffic Lights to decide if they
should launch a new business. More important, there are many ways of
successfully developing new businesses that do not involve disruptive
innovations.
A MORE COMPLETE THEORY
In this chapter we have compared the Traffic Lights with previous

theories about diversification. We have focused in particular on
Parenting Theory, not only because it has become the dominant the-
ory of corporate-level strategy but also because it influenced our
research. We were focused on companies who had succeeded in or
were ambitious to break the rules of Parenting Theory. We wanted to
find out if it needed some embellishment.
Our conclusion is that it does. The Traffic Lights point to a num-
ber of variables that are not given sufficient attention in Parenting
Theory or in any of its predecessors. As a result, the Traffic Lights
point to two additional logics for entering new businesses: saplings and
rare games (see Figure 6.5). This shows where saplings and rare games
would normally be plotted on the Ashridge Portfolio Display. The dot-
ted circles are meant to signal that only a few of the businesses
positioned in these parts of the display will be saplings or rare games.
In practice, for an Intel or McDonald’s there may be no saplings or rare
games.
As a further test of the proposition that the Traffic Lights offer a
more complete theory, we have compared the Traffic Lights with
176
THE GROWTH GAMBLE
Figure 6.5 Four strategies for new businesses
Christensen’s disruptive technology ideas. Again, we were able to

demonstrate that the Traffic Lights offer a fuller explanation of the
wide range of new business successes. A management team considering
new businesses will find more that pass the Traffic Lights than pass
Christensen’s disruptive innovation criteria.
177
CHAPTER 7
SEARCHING FOR
NEW BUSINESSES
A
ll companies face the question of what they should do to
increase the chances of successfully creating new businesses.
Managers want to know what organizational units and
process to set up. They also want to know what activities should be part
of preparing the ground and what activities will be needed once a
promising new business has been spotted. This latter question, what to
do once a promising new business opportunity has passed the Traffic
Lights, will be addressed in Chapter 9. In this chapter we deal with the
first issue: what a company should do to increase the chances of find-
ing a significant new business opportunity.
This question is complicated by the fact that we are focusing on a
particular kind of situation: companies, like Intel and McDonald’s, that
want to find some new business to solve the problem of a slowdown in
the growth of their core. In other words, we are trying to help compa-
nies that want to look outside their current core. In the previous chap-
ter we have referred to the current core as the heartland, and the focus
of this chapter is to advise companies what they need to do to increase
their chances of finding edge-of-heartland, sapling, or rare-game
opportunities.
Some authors argue that all companies should be investing a small
amount every year in new businesses. In this way they will have a
healthy “third horizon”: projects that may have a significant impact on
the company in five or ten years’ time.1 Other authors argue that com-
panies should not distinguish between innovation in their existing
businesses and new business creation. By having a healthy innovation
THE GROWTH GAMBLE
process they will create both new products and new businesses.2 Other
authors are more aggressive. They argue that every company needs to
create a process for developing new businesses that is managed in par-
allel with the process for running existing businesses.3 Still others assert
that every company has a natural flow of initiatives driven by the
exploratory behavior of managers at all levels and that the only
requirement is to manage this flow effectively.4 Others propose that
companies should focus on disruptive technology developments, par-
ticularly those that are undermining the core businesses.5 Finally, some
argue that companies should stick to their core businesses and support
only those new business opportunities that are natural adjacencies.6
Our view is both different from and an amalgam of these ideas.
One of the difficulties of thinking clearly about new businesses is
that it is easy to become confused between efforts to develop new
products for existing businesses, efforts to add businesses in the exist-
ing heartland, and efforts to find edge-of-heartland opportunities,
saplings, or rare games. Investments in a central research function, for
example at Unilever or Intel, are often thought of as an appropriate
way of developing new businesses. However, most of the work of these
research functions will be and should be devoted to technologies and
products for the existing businesses and the existing heartland. So how
much should be invested in more speculative research that might lead
to new businesses outside the existing heartland? A similar question
can be asked of the new product development function, the corporate
venturing unit (if one exists), and the efforts of the strategic planning
team. The most obvious answer is that more should be invested in
exploring outside the heartland the more urgent the need. If growth in
the core is slowing fast, the company should pump up its investment in
research, new products, new ventures, and strategic analysis outside the
heartland. Nevertheless, our research has led us to be wary of this
seemingly obvious answer.
179
SEARCHING FOR NEW BUSINESSES
THE NATURAL FLOW OF NEW IDEAS
Every company has a natural flow of new ideas for new products, new
markets, and new businesses. This flow is dependent on the environ-
ment the company is in and the business model of the parent company.
The environment has an important impact. In the late 1990s, the
internet boom and the entrepreneurial bubble that it unleashed gave
rise to a period when many new businesses were being developed.
Almost every company was inundated with proposals. For example,
when we were doing research on the impact of the internet on corpo-
rate centers, one question we asked was how the corporate center han-
dled the flow of potential new businesses, joint ventures, and alliances.
A senior executive at AstraZeneca, a pharmaceutical company,
responded mischievously: “We have set up machine guns in the lobby!”
In order to deter the flow, managers were turning away every proposal
that did not have some privileged route of access.
Another example was tobacco company BAT. To respond to the
internal and external pressures, BAT set up two units. Imagine was set
up to work with managers or external agents to refine and develop new
ideas. Evolution was set up as an investment unit, to provide funds for
new ideas that seemed promising. While, with hindsight, BAT proba-
bly overreacted and AstraZeneca kept its feet more firmly on the
ground, both companies experienced a huge increase in the natural
flow of new business ideas.
Some industries, at certain points in time, have a higher rate of flow
than others. Industries faced with disruptive technologies, as most were
during the internet boom, have a higher flow. Industries undergoing other
major changes, such as changes in legislation (utilities), fragmentation
(financial services), or the collapse of a dominant competitor (computers
in 1990s), also have a higher flow. Industries that have been stable for
many years, where few of the sources of advantage are changing (mass-
market consumables, mining), have a lower flow of new business ideas.
The second factor affecting the natural flow is the business model
of the corporate parent company (and/or of divisions within the parent
company). Some companies, such as Philips or Canon, invest heavily
180
THE GROWTH GAMBLE
in research as part of the strategy for their core businesses. Inevitably,

even if the research is focused tightly on the needs of existing busi-
nesses, it will spawn new technologies and new product ideas that can
be developed into new businesses.
Other companies, such as 3M, have businesses built on innovation
as a source of competitive advantage. All of 3M’s businesses seek to be
the technology leader in their markets and to earn superior margins as
a result. This encourages managers throughout the organization to
search for technical or innovative solutions. Inevitably, these compa-
nies will also have a higher flow of new business ideas.
Other companies actively seek new businesses. Virgin is famous for
its openness to new ideas and entrepreneurial behavior. As a result, the
central team at Virgin receives tens of new business ideas every day.
When Richard Branson, the founder, flies on a long-haul flight, peo-
ple come up to him on the aeroplane with ideas for new businesses.
“After a transatlantic flight, he will come back with five to ten business
proposals,” commented Virgin’s head of strategy.
Virgin is therefore more like a venture capital company than a nor-
mal corporation. Richard Branson has demonstrated that he is pre-
pared to consider any interesting new business proposal. As a result, he
attracts a larger number of proposals than most companies would be
interested in. Virgin is not in an environment that is particularly
vibrant. The reason it has such a high flow is because the business
model of the parent company is more about spotting and launching
new businesses than administering its existing businesses.
In Chapter 1 we described the growth efforts at Intel and
McDonald’s. Without analyzing their situations in detail, we might
speculate that Intel’s natural flow of new business ideas is richer than
McDonald’s. This would not be because they have significantly differ-
ent corporate business models: both companies are heavily committed
to their existing businesses, and were exploring new businesses in an
attempt to find growth. Intel’s flow would be likely to be greater
because Intel’s environment has more new business activity.
Nevertheless, Mats Lederhausen of McDonald’s pointed out that there
are many new ideas even at McDonald’s:
181
We have plenty of people out there with ideas. What we do not have
are managers skilled at helping to accelerate these ideas. My team’s
main role is to act in support of the businesses to help them make more
out of the ideas they already have.
Robert Burgelman, a professor at Stanford Business School, has

devoted much of his research to understanding this natural flow of new
ideas.7 He recognizes that there are two sources: the top-down strate-
gic planning process and the bottom-up, autonomous initiative
process. The top-down process identifies changes in the environment,
proposes new solutions, and is stimulated by ideas from investment
bankers and alliance partners. The bottom-up process stems from
marketplace and technical initiatives taken by managers in response to
customer requests, partner initiatives, and perceived opportunities.
Often these bottom-up projects start out as “skunk works,” below the
radar of top-down plans. However, if successful, they get written into
the plans retrospectively. Burgelman argues that the bottom-up
process is as natural a managerial activity as top-down planning: it
should be recognized as a normal part of corporate development. We
agree.
SHOULD COMPANIES TRY TO INCREASE THEIR NATURAL

FLOW?
Managers often conclude that they are not getting enough new ideas.
They look at companies like Virgin or 3M and envy their higher flow
rate. Since they can do little about the environment, they try to solve
the problem by changing the corporate business model to encourage
more ideas.
Typically this will involve setting up a corporate venturing unit or
business development function. It will involve analyzing trends in the
industry to generate more ideas. It may involve increasing investment
in research or new product development. It may involve a series of
workshops to generate ideas and stimulate the thinking of managers at
all levels in the organization. These workshops or brainstorming
182
THE GROWTH GAMBLE
sessions are often supported by corporation-wide innovation training.

It may involve setting up a small M&A team to look at acquisition tar-
gets. It may involve creating multiple sources of finance for new ideas
in order to encourage more internal entrepreneurs to come forward.8
These are the sorts of things that both Intel and McDonald’s did in the
1990s to increase their chances of finding new businesses.
Unfortunately, trying to pump up the flow of new ideas does not seem
to work. More new ideas are generated and processed, but the flow of
successful, significant new business initiatives does not seem to
increase.
Managers should use the New Businesses Traffic Lights to screen
ideas that are part of the current flow and invest only in those that get
sufficient green lights. Sometimes this will lead to a large number of
potential new investments. If so, the company needs to be more selec-
tive until it has a list of projects that matches its managerial capacity.
Sometimes there will be few or no investment opportunities. When
this occurs, managers should focus on maximizing the performance of
their existing businesses (see Chapter 4) rather than launching initia-
tives to pump up the flow of ideas. Certainly, managers should remove
blocks in their existing cultures or processes, and they may need to
increase the innovative activity within their core businesses. But they
should not view a lack of good opportunities as a failure that needs to
be corrected. In our experience it is the norm rather than the exception
for many companies.
We recognize that there are pressures on managers to grow and that
these pressures often demand some investment in new businesses. The
environment may be bursting with new proposals, managers’ existing
businesses may be being undermined by new technologies or other
changes, or their current businesses and their current corporate busi-
ness model, though sound, may not be producing enough growth.
Rather than viewing these pressures as a call to arms, managers should
treat them as pitfalls to be handled cautiously. By auditing their natu-
ral flow of ideas, managers can assess whether there are plenty of
opportunities or only a few. By clinging to the New Businesses Traffic
Lights, managers can resist the temptation to roll the dice when the
183
chances of success are low. By having a bias in favor of focusing on the

core, managers can ensure they do not prematurely lose faith in or
focus on their core businesses.
Of the 50 or so examples in our database of successes, none started
out as an initiative stimulated by a desire to “get more ideas into the
hopper.” They all occurred through the natural flow processes. They
either emerged from below as a result of initiatives taken by managers
near the marketplace or they were part of top-down strategizing. Some
also came as unexpected and unplanned opportunities brought to the
company from outside. In other words, the successes came from the
natural processes, not from forced processes.
Johnson Matthey’s move into manufacturing controlled substances
is an example. Johnson Matthey, an expert in platinum, had worked
with Bristol Myers Squibb to pioneer the use of platinum in anti-
cancer treatments. The platinum for these treatments was manufac-
tured in a plant in West Deptford, New Jersey.
Looking for growth opportunities, the company considered other
opportunities to supply the pharmaceutical industry. After some care-
ful analysis, management decided to apply for a license from the Drug
Enforcement Agency to manufacture controlled substances. The com-
pany’s experience of managing a secure facility and producing small
volumes using difficult catalysis proved decisive. Today, Johnson
Matthey is the number three supplier of analgesics in the US.
Nevertheless, managers argue: “But we want to be more like 3M
and Virgin. If we had a corporate-level business model that champi-
oned new businesses and attracted internal and external entrepreneurs,
we would be able to have a continuous flow of new businesses like
them.” And this is true. The problem is the size of change that is
needed. In order for a McDonald’s or an Intel to operate a corporate-
level business model like 3M or Virgin, these companies would have to
change too many things. They would have to change most of their top
managers, certainly including their CEO. They would have to change
many of their most cherished management processes. They would
have to change their portfolio of businesses, or at least signal widely
that the management team in charge of their core business has no
184
THE GROWTH GAMBLE
privileges over the teams of lesser businesses. They would have to

change many of their board members. They would need to change
many shareholders.
Interestingly, this objective—to become more like 3M or Virgin—
is the advice given by a large percentage of the current literature.9
Companies are encouraged to develop new corporate processes and
make other changes in order to be more attractive to new business
ideas. Not surprisingly the success rate is low, if not non-existent. Even
some of the strongest proponents of this kind of change admit that no
companies have yet succeeded in doing what is necessary. The reason
is that companies are organisms with well-entrenched habits and
processes. Improvements are possible, but dramatic change is not, at
least not without dramatic change in personnel. Moreover, even if dra-
matic change were possible, it might not be appropriate. Would Intel
run its microprocessor business as well if its corporate team operated
more like the corporate team at 3M or Virgin? The answer is no. The
likely cost in loss of performance in the core business would be a mul-
tiple of any gains made in new businesses.
By pouring cold water on efforts to generate a richer flow of new
business ideas, we are not trying to discourage companies from inno-
vation or entrepreneurial activity. We are trying to discourage them
from investing a lot of energy in forcing the pace.
Gary Hamel and his colleagues at Strategos argue that companies
should focus their innovation efforts on regenerating core businesses.
Once their core businesses are effectively using innovation to improve
performance, there will be a knock-on effect for new businesses: some
of the ideas will form part of the natural flow of new business ideas. We
have not seen this happen, but it makes sense. Innovation in existing
businesses is aimed at something managers understand, where their
instincts about what makes sense are likely to be sound. In these cir-
cumstances, managers have a feel for how much to spend and when
ideas are becoming foolish. New business ideas that emerge from this
process are likely to be sensible.
However, when the innovation effort is focused on areas where
managers have little understanding, their ability to know how much to
185
invest in the search effort or when ideas are losing touch with reality is
curtailed. As a result, little of value comes out.
For example, in an attempt to stimulate growth, the Prudential
identified the ageing population as an interesting trend. Clearly this
trend was having an impact on its core insurance businesses. But the
question managers asked was whether the Prudential could exploit this
trend outside the insurance industry. To pursue this idea, it hired a
major consulting company to help look at the opportunities. Obvious
ideas like care homes were considered, along with less obvious ideas
like housing or holidays in warmer climates like Spain. Unsurprisingly,
nothing emerged that looked attractive to the Prudential, and the
effort was closed down after quite a bit of money had been spent. We
believe that companies can avoid these kinds of wasteful exercises by
relaxing into the natural flow of ideas that will emerge anyway.
WHY FORCING THE PACE OF IDEA GENERATION DOES

NOT WORK
When we suggest to managers that they should “relax into the flow,” we
get odd looks. Are we smoking something? Surely, managers argue, it is
possible and sensible to increase the number of new business ideas that
are being considered: to fill the hopper with ideas? It seems such an
obvious thing to do, and it is more managerially attractive than waiting
for something to turn up. Yet we are arguing that it is “make work”:
work that makes managers feel good, but fails to add anything useful.
Our argument is simple. It concerns people. Successful new businesses
depend on the leadership of rather unusual managers (our leader/sponsor
Traffic Light). These managers normally have an insight into some aspect
of the market or business model that enables them to see opportunity
where others do not. But not only do they have the insight, they also have
the entrepreneurial initiative to get the new business going.
When managers with the entrepreneurial courage have important
insights, they are only too keen to tell other managers about the
opportunity and to start lobbying for support from their organization.
They will try to influence the top-down strategy process or create a
186
THE GROWTH GAMBLE
bottom-up skunk works. In other words, when the natural flow of

things brings together special insight and entrepreneurial talent, you
are not going to have to look very hard to find it. It will not be neces-
sary to set up search parties to turn over every stone. It is more likely
that the manager with the idea will beat your door down whether you
are looking or not. At Johnson Matthey, Forrest Sheffey, the CEO of
the pharmaceutical materials division, championed the move into con-
trolled substances based on his intimate knowledge of the sector.
If instead of waiting for the natural flow to produce ideas managers
try to force the pace, plenty of ideas can be generated. In a recent dis-
cussion, one consultant proudly explained that the idea-generation
process he had been leading, involving innovation workshops, sector
surveys, and incentives, had produced 387 ideas for new businesses for
the client. The problem is that 99% of these ideas will be lacking a key
ingredient: an entrepreneurial manager willing to lead the new busi-
ness who has a significant insight about the market or the business
model and sufficient status and support within the parent company.
When properly screened, the client will end up with the same number
of promising projects as would have been identified if the work had
just focused on the ideas that were available before the consulting
project was launched. What is more, with 387 ideas the temptation is
to ensure that 5% or 10% pass the screening.10 This means the client
will end up with more projects and more investment but no more
successes.
“But this contradicts the normal behavior within a business,” some
managers argue. “If you are losing market share, one of the tried-and-
tested ways of responding is to pump up your new product develop-
ment process. Why would this not work at the portfolio level?” The
answer links back to the people argument. In an existing business,
managers understand the market, technical risks, and their own skills
well enough to be wise at selecting new products with a good chance
of success. When screening new businesses involving some unfamiliar
markets, technologies, and business models, managers are less able to
distinguish the good ideas from the bad ones. They are therefore eas-
ily tempted into pursuing ideas that will not pass the Traffic Lights.
187
WHAT IMPROVEMENTS CAN BE MADE TO THE

NATURAL FLOW?
The message of this chapter so far has been that managers should not
try to pump up the natural flow of ideas. The argument is that it
achieves nothing in the short term. But is there not something a com-
pany can do over the longer term? The answer is yes, although not as
much as many authors suggest. The next two sections look at the top-
down and bottom-up processes and describe the type of action we
believe most managers should take.
IMPROVING TOP-DOWN PROCESSES
While we believe in discouraging managers from trying to force the

pace of idea generation, we are not trying to discourage them from ana-
lyzing their situation and thinking hard about their corporate strategy.
The most important work that top managers can do is to produce a
vision of how they are going to develop their portfolio of businesses and
articulate why this vision is likely to create value for all stakeholders.
Top-down, corporate-level strategy involves a number of steps. First,
managers will assess the industries in which they are currently compet-
ing. This work ought to have been done in the business-level plans, but
it often needs extending. Particular attention needs to be given to the
periphery of the industry, where disruptive innovations may be evident
or where new business opportunities may be emerging (see Box 7.1).
The overall objective of the analysis is to make sure that managers
have thoroughly explored the developments around their existing
businesses.
Second, managers will allocate each business to one of four cate-
gories—develop, improve, maintain, or harvest—based on the busi-
ness’s longer-term potential in its industry. Though simplistic, this is a
good starting point for achieving focus. It also stimulates challenging
debate with business managers.
Third, managers will develop a list of corporate-level skills and
resources. These will then be used to screen other industries to see
188
THE GROWTH GAMBLE
Box 7.1 Analyzing the periphery
Strategy work usually focuses on the main competitors and the core of the market.New
business models,however,often emerge from the edges of existing businesses and may
go unnoticed unless special attention is given to analyzing the periphery.There are three
steps.
1 First, managers should record all the competitors and business models that are in
some way connected to the current business.These may be competitors up or down
the value chain, in different channels, in adjacent segments, in substitute technolo-
gies, and so on. Particular attention should be given to new competitors and small
but successful competitors.
2 Second, some of these competitors should be analyzed in more detail—reverse
engineered to understand the economics of their business models. Choosing which
ones to analyze is an art.The objective is to focus on those competitors from which
the company has most to learn.
3 The final step is to extract the messages from these analyses.The messages may be
about potential disruptive technologies or business models.They may be about new
market segments and new customer expectations.They may be about innovations
in any aspect of the business.It is important that this assessment is not done solely
by the senior managers of the existing businesses.Their commitment to the exist-
ing business model may make them blind to changes that are happening in their
industries.The team extracting the messages should be broadly based and include
managers from the periphery, whose visceral understanding of the trends may be
better than that of managers at the center.
where else these skills could be deployed (see Box 7.2). These other
industries or markets should include those adjacent to the ones in
which the company is already competing. However, the analysis should
not be limited to adjacent industries. The industries chosen will
depend on the corporate-level skill.
The final step is to decide which parenting opportunities and
which parenting propositions the corporate level currently is or could
become good at. This is done by deciding which parenting opportu-
nities offer the biggest value increment and which best match the
189
Box 7.2 Defining corporate-level skills
The skills and resources at the parent company level are an important input to
corporate-level strategy.11 There are four steps involved.
1 Managers should identify where the existing businesses need most help from par-
ent managers. By identifying the main tasks of each business and predicting which
of these tasks managers will find hardest to complete well, parent managers can
identify the opportunities to help (parenting opportunities).
2 Managers should record the mistakes that are typically made by managers in the
industries the company is in.The analysis should include mistakes the company has
made and mistakes made by competitors.Mistakes include things like investing too
much at the top of the cycle, unnecessarily creating a price war, underinvesting in IT
or technology,and so on.Having developed a list of mistakes,managers should then
examine each and assess whether the parent has been good at helping businesses
avoid these mistakes or whether it has been complicit in the mistakes.
3 Managers should record the different ways in which corporate functions and line man-
agers are currently trying to help and influence the businesses.These “parenting propo-
sitions”will be based on the views of the current managers at the corporate level.
4 Managers should record the habits,biases,and normal instincts of the parent managers
as a group. It is useful to ask questions about how the parent normally behaves and
why.What is their usual reaction to a threatening competitor? How do they typically
deal with underperformance? How do they decide what is a suitable growth rate?
These four questions provide data for a strengths and weaknesses analysis of the corpo-
rate levels.
skills of corporate-level managers. The objective is to build around

some corporate-level skills where the company has or can reasonably
expect to be able to create an advantage over competing parent
companies.
Finally, managers can develop a corporate-level strategy that
defines which existing businesses and which new businesses best fit
with corporate-level skills. The strategy also defines how the
corporate-level plans will help existing businesses and new businesses
maximize their performance.
190
THE GROWTH GAMBLE
Whitbread’s work on corporate strategy in 2002 is a good example

of effective top-down strategizing. The company has a diverse port-
folio of restaurant and leisure brands, mainly in the UK. Some of these
brands—Marriott Hotels, Pizza Hut, and TGI Friday’s—are managed
in collaboration with their US brand owners. Others, such as Beefeater
(steak houses), Travel Inn (low-cost hotels), David Lloyd Leisure (ten-
nis clubs), and Brewer’s Fayre (pub restaurants), are UK-focused
brands.
Whitbread’s top-down review examined each brand and its market
sector. Travel Inn was allocated to the “develop” category and
Beefeater to the “improve” category. Travel Inn was growing fast in the
UK and had clear opportunities to grow overseas. Beefeater was the
UK’s leading steak restaurant, but was underperforming. The priority
was to ensure the brand was earning its cost of capital.
In addition, the work involved surveying 30 other market sectors.
Changes in the way consumers used their leisure time, such as the
increasing link between eating-out occasions and a leisure activity,
were creating new opportunities.
The 30 market sectors were put through a brief screen and quickly
reduced to two that appeared to offer promise for Whitbread.
Whitbread defined its corporate-level skills as developing leisure
brands, driving performance, improving brand positioning, and devel-
oping a pool of managers good at operating leisure brands. While
these skills did not point especially at any sub-sector in the leisure
industry, they fit well with the two opportunities identified.
One of these two interesting sectors, an opportunity closely linked
to an existing business, was worked up into a business plan. Within
nine months of the strategy review, the business proposal had been
approved and work had started on the first site.
Another part of the strategy looked at the potential for expanding
two existing brands into new geographic markets. For one brand the
first step was a joint venture in Spain. For the other brand the way for-
ward was an extension of its franchising concept to other markets.
Whitbread’s corporate vision continued to be focused on leisure and
restaurant brands. However, in the medium term the plan now
191
included an additional UK leisure brand and an ambition to develop

some brands internationally.
We like the Whitbread story because the top-down strategizing
resulted in a few crisp decisions, because it took less than a year,
because the new business efforts were seen as an integral part of the
corporate strategy, and because managers across the group were com-
mitted to the initiatives.
Despite the advice of some authors that companies should do less
top-down portfolio strategizing and more bottom-up experiments, we
believe that crisp, top-down thinking is evident in all but a few of the
companies in our success database. Without it, companies easily fall
into the trap of investing in a range of new initiatives, none of which
has sufficient corporate support to ensure success.
IMPROVING BOTTOM-UP PROCESSES
Since the most important insights about markets and business models
normally occur first to managers close to the market or close to oper-
ations, it is not surprising that every company has a natural flow of
bottom-up ideas and ventures. How can this bottom-up process be
effectively facilitated?
At one extreme, there is a danger that the company is so focused and
so controlling that it does not tolerate any activity outside its existing
businesses. Entrepreneurial managers in these situations take their
ideas elsewhere.
At the other extreme, companies can create so much slack and tol-
erate so many promising new initiatives that every dreamer is funded
and managers are distracted from driving the core businesses forward.
Most authors suggest that companies are too focused and control-
ling. They argue that, until companies change their cultures to give
more support to managers ready to take a risk, no new business ideas
will emerge. Clearly, some companies are like this. However, our
observation is the opposite. We believe that most companies are too
tolerant of new ventures and new projects: they support too many
improbable initiatives and take too long before cutting the funding to
192
THE GROWTH GAMBLE
failing projects. Hence our bias is to encourage companies to recognize

the existence of the bottom-up process, but to improve with tough love
rather than open arms.
We should, however, admit that we have not researched bottom-up
processes in detail. Our views are based on experience, on the shadow-
ing work, and on the database of successes. Fortunately, Robert
Burgelman has studied the bottom-up processes rather carefully, in
particular in his work with Intel. Although we do not agree with all of
his recommendations, much of his advice, which we summarize below,
is sound.12
1 Promote a “rugged, confrontational/collegial culture.”13 The company

needs a way for bottom-up thinking to challenge top-down think-
ing. The best solution is a process of debating issues that is vigor-
ous, issue driven, and indifferent of rank. It is also a process where
disagreements do not interfere with execution. Intel, according to
Burgelman, has both features. “Constructive confrontation” is the
Intel name given to robust debating, and “disagree and commit” the
label for the execution follow-through once the confrontation is
concluded.
2 Tolerate sponsored initiatives. Encourage the CEO to explicitly toler-
ate a few initiatives that do not easily fit into the current corporate
strategy, but are sponsored by the divisions reporting to the CEO.
3 Embrace ambiguity. Suspend, for new ventures, the drive for cer-
tainty that is often part of the budgeting and planning process. This
may require more frequent reviews and milestone reporting, but
should not result in interference.
4 Don’t separate. Those involved in bottom-up venturing should not
be isolated from the mainstream. Linkages are important to the
success of the venture, to the willingness of the parent to keep
supporting it, and to the process of developing the corporate
strategy.
5 Bolster the will to terminate. Bottom-up initiatives are experiments
and should be terminated if they are not making progress. Top man-
agement needs an explicit termination process.14
193
6 Ensure sufficient general managers. New ventures require managers

capable of linking new ideas with the marketplace and accessing
resources from different parts of the company. Without a sufficient
number of managers with generalist skills, ventures have less chance
of succeeding.
As opposed to authors like Rosabeth Kanter, Burgelman does not sug-

gest radical changes. He acknowledges that there is a bottom-up
process that exists regardless of the context. However, he believes that
there are actions managers can take to facilitate the process. We agree
with this positioning.
One area where we disagree with Burgelman is his presumption that
the bottom-up process will be and should be continuously creating new
ventures. We believe that, at certain points in time, the bottom-up
process may be generating no new ventures at all. While it is not diffi-
cult to resolve the issue by observation, and indeed some companies at
some points in time have no bottom-up-generated ventures,
Burgelman’s view is that this is a signal of failure. In contrast, our view
is that a lack of bottom-up ventures may be totally appropriate. It may
be caused by a lack of opportunities or the existence of more important
tasks, such as driving the existing businesses. So managers worried
about the vibrancy of their bottom-up processes need to analyze the
situation carefully before they make too many changes.
An example of careful analysis and thoughtful change is IBM’s
Emerging Business Opportunities (EBO) process.15 Faced with low
growth, a concern that IBM had missed out on some of the emerging
opportunities in the computer and internet industry, and a belief that
insufficient new ventures were coming through the bottom-up process,
CEO Sam Palmisano commissioned a group of managers to suggest a
process for boosting IBM’s success in new businesses.
The team concluded that there was a problem: IBM’s core manage-
ment processes and complex organization matrix were blocks to the
bottom-up development of new businesses. The management
processes were focused on profitability rather than value, and the
matrix structure required that new businesses win support separately
194
THE GROWTH GAMBLE
from research, sales, and a business group. After some consideration,

the team decided to set up a program to support the bottom-up
processes. First, some new businesses emerging in the divisions were
selected as EBOs. They were given some protection from the normal
resource-allocation process, some additional funding from a corporate
resource pool, and a good deal of attention and help from central
research and strategy.
Over the first three years the program was considered a success. It
gave additional attention and support to some new businesses that
might otherwise have made little progress. The program also encour-
aged some managers to propose businesses as EBOs when under dif-
ferent circumstances they might have focused their energies only on
the existing businesses. By 2003, the EBO businesses were generat-
ing sales of $10 billion (although reputedly not much profit). Some
of these initiatives would clearly have happened in any case, but the
program was believed to have given these projects a substantial leg
up.
IBM’s growth challenge remained, however. The EBOs were
adding to growth, but not enough to make a significant difference. In
2003, IBM decided to try to add significantly to its current program of
18 EBOs. This involved graduating some of the existing EBOs and
“filling the hopper” with a larger number of new EBOs.
If this expansion happens (which is described in a Harvard Business
School case but somewhat contradicted by internal sources), we expect
IBM’s EBO program to become less successful rather than more. The
desire to use the program to “fill the growth gap” is in our view the
wrong motivation. The program should be used to help reduce the
barriers to the natural flow of new business projects rather than pump
up the flow in order to hit a growth ambition. Identifying the right bal-
ance of initiatives is hard, even in a company as thoughtful as IBM.
This is especially true against the backdrop of the pressures to grow
that influence all managers.
195
BRINGING THE TOP-DOWN AND BOTTOM-UP PROCESSES

TOGETHER
With a crisp top-down process and a blockage-free bottom-up process,

there is an issue of how to bring them both together. This is another
area where our thinking diverges from Burgelman’s. He, like other
authors, argues that managing both the bottom-up process and the
top-down process is very difficult:
Top management must also help the company exercise both disciplines
simultaneously. Companies find it extremely difficult to do so.16
Burgelman believes it is difficult to manage the two processes because

the bottom-up process is intuitive and entrepreneurial, while most top-
down processes are locked into past concepts of the company’s corpo-
rate strategy.
We think that both of these problems can be overcome. We see the
bottom-up process as a healthy entrepreneurial activity that needs to
be kept in proportion. The way to do this is to impose on the entre-
preneurial managers a clear set of criteria for deciding which bottom-
up ideas will be supported and which will be rejected. The Traffic
Lights provide the criteria. So long as the Traffic Lights are fully com-
municated down the line, managers at any level can experiment and
launch new initiatives, either as skunk works or with approval from
middle managers, knowing whether their ideas are likely to be well
received or badly received higher up. The entrepreneurs can control
themselves.
Inevitably, some strong-minded managers will push forward ideas
that do not pass the Traffic Lights, or will judge some of the lights as
green when a more objective analysis would score them yellow or red.
When this happens, corporate-level managers should not hesitate to
use the Traffic Lights to expose the inappropriate venture. It can then
be formally closed down with a clear explanation of why. Other entre-
preneurially minded managers will not be discouraged unless they have
ideas that will be hard to justify using the Traffic Lights. We see the
196
THE GROWTH GAMBLE
corporate-level managers playing a critical role of disciplining the

entrepreneurial energies lower down.
Burgelman would prefer corporate-level managers to suspend judg-
ment until these bottom-up ventures have had a chance to develop fur-
ther, because it is hard for them to be sufficiently open-minded given
their commitment to the past corporate strategy. For example, it took
a change in leadership at Intel, from Moore to Grove, before the strat-
egy officially changed from memories to microprocessors. We recog-
nize this problem. But we think that the Traffic Lights provide the
solution.
The Traffic Lights are a generic set of criteria for screening new
business ideas. They are not dependent on the past corporate-level
strategy of any particular company. Inevitably, managers with different
views of where advantage lies and what the organization’s skills are will
score the lights slightly differently. But they provide a language that
top managers can use to communicate with front-line managers and
vice versa.
In some companies, such as Whitbread, the top-down strategy
development process defines clear paths for future development that
make sense when considered against the Traffic Lights. Once these
have been selected, managers lower down should be encouraged to
channel their entrepreneurial activity along these paths. However, a
manager lower down who has an idea that does not fit these paths can
still use the Traffic Lights to argue his or her case, and corporate-level
managers should give the proposal due consideration. In other words,
the Traffic Lights provide a language around which the top-down and
bottom-up processes can integrate.
Problems emerge when the top-down process creates confusion
lower in the organization. This happens when managers at either level
use inappropriate criteria to screen ideas, when top managers are
locked into outdated strategies, or when top-down screening judg-
ments are not consistent. Our research and Burgelman’s work show
that this kind of confusion happens frequently. Burgelman suggests
that the solution is achieved by middle managers acting as go-
betweens. Middle managers interpret corporate-level strategy for
197
managers lower down, while giving air cover from the inappropriate
thoughts of senior managers. At the same time, they interpret new ven-
tures for their bosses, helping them to see how they can be incorporated
in a redefined corporate-level strategy. In our view this puts an unrea-
sonable burden on middle managers. Moreover, we do not need to make
them into go-betweens. Disciplined use of the Traffic Lights can provide
the clarity that solves the problem: managers lower down know how
managers higher up are going to evaluate their projects and can use the
Traffic Lights to argue for adjustments to the corporate strategy.
Burgelman presumes “bias” among top managers and sees the
bottom-up process as an essential antidote. While this is an accurate
description of many companies, and Intel in particular, there are
enough examples of a better top-down process to demonstrate that
proper use of clear criteria helps top managers be wise and clears up
confusion lower down.
HOW MUCH SHOULD BE INVESTED IN NEW

BUSINESSES?
Our analysis of the top-down and bottom-up processes should help

managers decide what to do when they feel the need to create new
businesses. But it has still not answered the question: “How much
should we invest in new businesses?”
The simple answer is “enough to support all the new business ideas
that pass the Traffic Lights, constrained only by managerial capacity.”
In other words, if no ideas pass the Traffic Lights, managers should
make no investments in new businesses and use spare cash to buy back
shares or pay larger than usual dividends. If many ideas pass the Traffic
Lights, managers should invest in as many as the managerial capacity
allows. In a large divisionalized company this could mean investing in
10 or 15 new business projects, assuming two projects led by each of
five divisions plus some additional projects led by the corporate center
and a large business unit. This is about the number of projects in IBM’s
original EBO program.
198
THE GROWTH GAMBLE
It is also the number of new business projects on which Philips

chose to focus in the early 1990s. At a time when painful downsizing
targets were imposed on each division, managers decided to identify
those new business initiatives that were worth protecting from the cuts.
Thus 15 initiatives were singled out for corporate review and support,
and were put under the spotlight of the top 100 managers in the com-
pany. In total they absorbed at least 10% of R&D spending. They
included electronic lighting, flat-panel displays, cellular telephones,
medical ultrasound, high-definition television, interactive CD, and car
navigation systems. Over the next several years, a number were dis-
continued or divested. About half ended up as successful additions to
the business portfolio of their divisions.
Is there a financial limit on the amount that should be invested in
new businesses? In principle, no. For a new independent company,
100% of the capital is invested in “new businesses.” For large estab-
lished companies with growing core businesses, like McDonald’s or
Intel, the amount is likely to be less than 5% of total capital spend. For
a company that is exiting one core business and investing in a new area,
the amount invested in the new area may be close to 100%.
Managers, we have found, are not satisfied with our simple answer
to how much to invest. First, they believe that there ought to be some
minimum ongoing investment “in the future” or “in the third horizon”
or “to keep the pipeline full.” We hope that by this point in the book
we have persuaded you that, although this pipeline thinking is appro-
priate for new products and for research functions, it is not helpful
when companies are looking for new businesses.
The second argument is more difficult to respond to. Managers
argue that companies ought to invest more when their core business
is slowing down and less when it is growing fast. In addition, when
the industry is fragmenting or changing dramatically (as with finan-
cial services), managers should be investing more in new business
models than when the industry is consolidating (as with consumer
products).
It is possible to produce a simple matrix (Figure 7.1 overleaf) based
on these two dimensions: the degree of growth of existing businesses
199
Figure 7.1 Four types of industry environments
(high or moderate) and the nature of change in the industry (gradual

or disruptive).
The matrix divides the industry environment into four quadrants:
mature and sound, expanding, challenged by disruption, and prolifer-
ating and fragmenting. The matrix also highlights the very under-
standable pressures managers face in each of the four quadrants. Apart
from the benign top left quadrant, expanding, where there is continued
growth and little challenge from change, each of the other environ-
ments puts significant pressure on managers to develop new
businesses.
While we recognize the pressures managers are under, we do not
believe that they should let these pressures be the primary driver of
their strategy for new businesses. Interestingly, our database of suc-
cesses is spread across all four environments (Figure 7.2). Most were in
the bottom left-hand quadrant, mature and sound, perhaps because of
our bias toward major companies in mature markets. However, the
second-largest group was in the top left quadrant, expanding, and the
smallest in the proliferating and fragmenting quadrant. Hence man-
agers may have limited opportunities that pass the Traffic Lights when
200
THE GROWTH GAMBLE
Figure 7.2 Successes by industry environment
their industry is “challenged by disruption” and many opportunities

when their company is “expanding.” In other words, it is better for
managers to build their strategy for new businesses around those
opportunities that pass the Traffic Lights rather than around the pres-
sures they are under.
The industry environments matrix is useful in guiding managers to
give particular attention to different Traffic Lights at different times.
Intel probably sits in the top left quadrant. Its core business is sound
and expanding. Clearly, Intel should be considering opportunities adja-
cent to its microprocessor business. But managers should give special
attention to distraction risks, since any loss of attention on the core
microprocessor business will be very expensive. It may be appropriate
that the microprocessor business is a “creosote bush.”
In the bottom right quadrant, the industry is changing and creating
opportunities but existing businesses are under threat. This was prob-
ably what managers at McDonald’s thought they were facing in the late
1990s. The core hamburger business was being challenged by many
new fast-casual formats.
In this situation, the focus for managers should be to understand
these potentially disruptive challenges. For example, if the conclusion
201
is that the challenges will disrupt the hamburger business, then

McDonald’s would need to identify one or two of the new casual for-
mats to invest in. Once the choice has been made, McDonald’s should
be prepared to milk the core business in order to invest in the new for-
mat. Instead of worrying about distraction risks, managers in the bot-
tom right quadrant should be focused on making sure that they choose
a new area where they have sufficient advantage and where the profit
pool is not a dog.
Based on McDonald’s recent actions, it seems management has
decided that the threats from fast-casual businesses are not disruptive.
They are probably in the bottom left quadrant. As a result, their decision
to invest in new businesses only if they can become significant and with-
out any disruption to the core hamburger business seems appropriate.
DSM, now a Dutch life sciences company, has been a good bottom
right example, on two occasions. In the 1960s managers decided that
coalmining was no longer competitive in the Netherlands. This led
them to close the mines and invest in petrochemicals. By 1975, they
had completely exited the mining business and successfully entered the
chemical industry. In the 1980s, managers concluded that DSM would
be unlikely to be a winner in the consolidation of the petrochemical
industry. They backed a strategy of developing fine chemicals and bio-
chemicals. This led to some success in the life sciences industry. As
soon as these successes became clear, the company began to shed its
petrochemical businesses.
In some bottom right situations, none of the new business opportu-
nities pass the Traffic Lights. This happens when there are few, but
highly contested opportunities. The existing company does not have
advantage or the scramble for survival is making the profit pool into a
dog. The only rational choice is to focus on the existing businesses and
manage the decline gracefully. However, since managing a public com-
pany in decline is an unpopular activity with both managers and share-
holders, one strategic option is to sell the company to a management
buyout or private equity firm.
The bottom left quadrant is one where managers should be com-
fortable continuing to drive their core business, using the cash to buy
202
THE GROWTH GAMBLE
back shares or pay high dividends. These companies often have good
low-growth performance. New businesses should not be a priority, but
should not be ignored. The Traffic Lights analysis of opportunities will
give particular attention to distraction risks and leadership/sponsorship
issues. If sufficient lights are green, the company should be happy to
invest with confidence.
Anglo American’s move into European forest products is an exam-
ple. The existing businesses, primarily mining, precious metals, and
coal, were slow growth and stable. The company considered many
directions for expansion, but rejected most of them because it did not
have sufficient advantage or sufficient management talent. However,
when Tony Trahar became chief executive, he arrived with good
knowledge of the paper industry and of the management skills available
to expand this business. He therefore sponsored investments in the
European paper industry, placing the project in the hands of one of his
best managers. Over 10 years this business has become the most suc-
cessful forest products company in Europe.
Managers often presume that the amount they should invest in new
businesses should be determined by the degree to which their core
business is under pressure and the degree to which they have a pipeline
of new businesses in development. If the core is under threat or grow-
ing slowly and the pipeline is empty, the presumption is that they
should invest heavily in new ventures, acquisitions, and other
initiatives.
Our advice encourages managers to ignore the pressures they are
under and focus on the attractiveness of the opportunities that they
have in front of them. The nature of the pressures they face should
cause them to give more or less emphasis to different parts of the
Traffic Lights. However, the amount they invest should be deter-
mined only by the number of projects that get sufficient green
lights.
203
HOW TO ENSURE THE RIGHT PROCESSES FOR

NEW BUSINESSES
The following three guiding rules will help every company set up the
right processes for developing new businesses.
1 MANAGE THE FLOW OF NEW BUSINESS IDEAS
To manage the flow of new business ideas, start with the strategies of
the existing businesses, avoid overstimulating the idea flow, yet design
a process for capturing and screening ideas.
Time and again, we have found that failure to understand the
dynamics of the existing businesses leads managers to underestimate
their potential or the challenges they face. This can lead to over-
estimating the managerial and other resources available to support new
businesses and underestimating the distraction costs. Without a thor-
ough analysis of the existing businesses, managers do not have a sound
platform from which to approach the screening of new business
projects.
The strategy review should not only focus on the current business
model and current product–market segments, but also on adjacent
business models and on new disruptive competitors. This will give sen-
ior managers across the company the insights needed for developing
top-down ideas for new businesses. However, just as often the review
will conclude that there are sufficient management challenges and suf-
ficient growth opportunities within or around existing businesses.
Excessively stimulating the idea flow has many disadvantages. It
encourages managers to launch too many new ventures. It distracts
them from the one or two that have real potential. It causes the rest of
the organization to give minimal support because of the high failure
rate.
At the same time, senior managers need an organized way to cap-
ture bottom-up ideas that originate within the company, be it from
R&D or from marketing and sales. It is important people know that a
process exists to deal with such ideas and that decisions get taken in a
204
THE GROWTH GAMBLE
timely and rational manner. Some companies have used a venture

board to fill that role. Others simply use the top executive group to
deal with such decisions.
It is extremely helpful to managers in dealing with both the top-
down and bottom-up flow of ideas to be able to refer to a generally
accepted and clearly communicated strategic framework. People
mostly look to the CEO to spell out the company strategy in terms that
will help them understand which ideas might be welcome and why cer-
tain initiatives are supported while others are rejected.
2 DILIGENTLY USE THE NEW BUSINESSES TRAFFIC LIGHTS
Whether a company invests in new-ideas workshops or limits itself to

looking at opportunities that emerge, the Traffic Lights can help
ensure that foolish ideas do not get through the screening process. In
fact, if the Traffic Lights are widely communicated as a tool for assess-
ing the “strategic business case,” managers at all levels will feel empow-
ered both to propose ideas that fit and reject ideas that will not pass the
Traffic Lights.
Every tool can be used inappropriately. Managers who are deter-
mined to get their ideas approved can consciously or unconsciously
bias their thinking. Hence the Traffic Lights need to be applied as part
of a decision process that includes some skeptics and some managers
less committed to developing new businesses. The tool itself is not
enough. The company also needs a governance process that is objec-
tive and strong enough to turn down the pet projects of the CEO, the
CFO, or the head of the largest division.
3 EFFECTIVELY PARENT EACH NEW BUSINESS INITIATIVE
Once a new business idea passes the Traffic Lights, managers need to
pay considerable attention to parenting it appropriately.
Multidivisional companies will often have a number of these initiatives
ongoing, and will need a process for assigning a good sponsor, putting
them at the right place in the organization, orchestrating the right
205
amount of support, and reviewing progress effectively. The EBO pro-

gram at IBM and the new projects program at Philips are examples.
The toughest decision is faced when progress deviates from the
original plan in some major way. This will occur several times in the
course of developing a new business. The choice between reposition-
ing the initiative, altering its objectives, increasing the commitment of
resources, or pulling the plug often involves agonizing judgments. We
say more about how to address this issue and how to parent new busi-
nesses in Chapter 9.
206
CHAPTER 8
IS THERE A ROLE FOR

CORPORATE VENTURING?
I
n earlier chapters of this book we have argued against corporate
venturing units. We have given evidence that corporate ventur-
ing units do not help companies develop significant new busi-
nesses and we have explained why. In this chapter, we want to ask the
question: “Are there any uses for venturing techniques inside large
companies?” In answer, we will define five types of venturing unit that
can add significant value if used in the right circumstances—five dif-
ferent venturing business models. None of these five, however, will
help a company develop significant new legs.
It is worth starting with some history. In 2002, companies were
questioning the wisdom of their investments in corporate venturing
units for the third time in 40 years. These incubators, corporate ven-
ture units, and new business development units were set up predomi-
nantly in the late 1990s as vehicles for companies to achieve a variety
of objectives: to explore new business opportunities created by the
internet; to speed growth; to exploit internal resources; and to tap into
the entrepreneurial boom that was exploding all around them (Figure
8.1).
Many concluded that these investments were part of a managerial
madness connected to the internet bubble. Like Compaq, Vodafone,
and Royal Sun Alliance, they closed down their corporate venture unit
and resolved to avoid such foolishness in future. Some, like Diageo,
Ericsson, and Alcatel, continued in a cautious manner uncertain
whether the current bust would be as temporary as the previous boom.
A few, however, such as Intel, Johnson & Johnson, and Nokia,
IS THERE A ROLE FOR CORPORATE VENTURING?
Figure 8.1 Investment levels for corporate venture capital, 1980–2002
continued undeterred. They had a logic for their corporate venturing

that was still sound, and they remained committed. Finally, some com-
panies stepped up their commitment and a few, most notably Unilever,
established new vehicles to stimulate and invest in new ventures.
McDonald’s set up a venture unit in 2003, but in a rather different for-
mat from Unilever. While Unilever was using venturing to explore
new consumer trends, McDonald’s set up McDonald’s Ventures as a
way of containing its remaining new business investments.
The reason for this confusing picture is that corporate venture units
encompass five different objectives each requiring a different business
model.1 Harvest venturing involves setting up a unit to turn spare inter-
nal resources into cash. Ecosystem venturing involves investing in com-
panies that are complementary to existing businesses. They may be
suppliers or customers, or, as at Intel, companies making complemen-
tary products or software. The objective is to benefit existing busi-
nesses by creating a more vibrant environment for these businesses to
operate in. Innovation venturing involves using venturing techniques to
stimulate entrepreneurial activity within an existing function of an
208
THE GROWTH GAMBLE
Figure 8.2 Success at different types of venture units
existing business. Private equity venturing involves setting up a unit to

compete in the private equity and venture capital industry. New leg ven-
turing involves setting up a unit to develop internal or external projects
and businesses with a view to generating significant new revenue
streams (i.e., new legs) for the corporation. The first three of these
business models appear to be commercially sound when the appropri-
ate circumstances exist. Private equity venturing appears to have been
viable in the past, but is unlikely to be appropriate in the future except
for a very few companies. New leg venturing appears to be fundamen-
tally flawed regardless of the circumstances (see Figure 8.2).
The reasons for failures, both in the last three years and in the pre-
vious two corporate venturing booms (early 1970s and mid-1980s), are
the same:
❏ Some companies chose commercially sound business models but were

caught out by the boom and bust in valuations. They invested too
much at the top of the cycle. These companies should not abandon
their efforts now. Instead, they should develop processes to curb enthu-
siasm and limit investment budgets when valuations are booming.
209
❏ Some companies chose new leg venturing and reaped the conse-
quences, incurring huge losses. These companies need to under-
stand why their efforts failed, so that they can avoid making the
same mistakes in the future. They also need to recognize that some
forms of corporate venturing may still be appropriate for them.
❏ The vast majority of companies, however, had unsatisfactory results
because they failed to understand that different venturing objectives
require different business models. As a result, they set up corporate
venture units with broad remits and multiple objectives. Without
focus and an appropriate business model these units floundered,
often losing credibility with the rest of the organization before a
viable model could be developed. These companies need to reflect
on their experience, decide whether any of the viable business mod-
els is appropriate, and define a venturing strategy for the future that
is built on firmer foundations.
In this chapter we will describe the five types and identify the key man-
agerial challenges associated with each (see Box 8.1 on pages 228–9).
In doing this, we build explicitly on a number of earlier studies of cor-
porate venturing units.2
HARVEST VENTURING
The objective of turning excess corporate resources into cash is one

that does not always need to involve the techniques associated with
corporate venturing. For example, the corporate property function is
frequently involved in sub-letting spare office space and selling spare
land or developing it for sale. While the corporate property depart-
ment usually is a separate unit, it does not normally use venturing tech-
niques to stimulate or process projects.
However, for some corporate resources, such as technology, mana-
gerial skills, brands, and even fixed assets, a venturing approach can be
used alongside or in place of the licensing or selling of these assets. We
would describe any unit that is set up to “harvest” spare resources as a
210
THE GROWTH GAMBLE
harvesting unit whether it employed venturing techniques or not. For

example, BG Property was set up to exploit the large land holdings of
the BG Group, holdings that had been inherited from a time when the
company, then British Gas, was government owned. Venturing tech-
niques were a minor part of a broader harvesting business model that
was primarily driven by the need to prepare land for sale. Harvesting
is, therefore, a concept that is much bigger than harvest venturing.
However, in this chapter we are interested in units that employ ven-
turing techniques as an important, if not prime, part of their business
model.
Lucent New Ventures Group is a good example. It was set up in
1997 to commercialize intellectual property and technology that was
not immediately supported by a business unit. Bell Labs, the research
arm of Lucent, is famous for its innovations, patents, and technology.
Lucent NVG’s objective was to create value by exploiting these under-
used assets. Over four years, the unit evaluated over 300 opportunities,
started 35 ventures, and drew in $350 million of external venture
capital.
Tom Uhlman, who was in charge of Lucent NVG, attributed the
success to having financial return, rather than some strategic goal, as
the primary objective. Could Lucent NVG have succeeded if it was
linked more tightly to the parent company? Uhlman asserted:
No! We certainly would have failed to get outside investor interest, and
probably would have made some lousy decisions.
However, by 2001, the market for new ventures was weak, and Lucent
Corporation was so short of money that the New Ventures Group was
sold to a UK private equity firm, Coller Capital.3
BT Brightstar was set up in 1999 and had similar ambitions: to har-
vest value from the 14,000 patents and 2,500 unique inventions arising
at BT’s R&D establishment at Adastral Park and to effect cultural
change in what was still a rather traditional organization. BT
Brightstar operated a rigorous stage-gate process for progressing ven-
tures and enjoyed some early success. By 2001, 330 ideas had been
211
presented, with 4 transactions concluded, and revenues of £30 million

generated. As the dot-com bust took hold, BT Brightstar found the
harvest venturing model increasingly difficult to execute, primarily
because it was relying on external venture capital funding for its spin-
outs, and the market for such investments had dried up. In January
2003, BT sold a 60% stake in Brightstar, also to Coller Capital, as its
only means of gaining funding for its pipeline of start-up ventures.
A harvest venturing unit is an appropriate vehicle under three
circumstances:
1 Management must believe that the company has some resources—

intellectual property, brands, assets, and so on—that are not being
fully exploited.
2 New businesses need to be created around these resources in order
to exploit their full value. For example, a technical development
may not have any obvious value until its commercial potential has
been demonstrated. Unless a new business is needed, venturing
techniques are not necessary. The company can sell or license the
resource.
3 The third condition is that the resources are not needed either by
the existing businesses or as new growth platforms. If they are
needed for a non-harvesting purpose, one of the other venturing
business models may be more suitable.
In both BT and Lucent, these conditions existed in the late 1990s.

Managers were convinced that the research laboratories contained
valuable technologies. The technologies were not needed by the exist-
ing businesses. And the booming venture capital industry was hungry
for new businesses that contained promising technology. These com-
panies needed to set up harvest venturing units in order to exploit this
opportunity. As it turned out, both units found that the window of
opportunity—the period of venture industry hunger—was quite short-
lived, and their new business proposals had few other markets. The
subsequent buyout and rationalization of both units by Coller Capital
was therefore an appropriate response to the new market conditions.
212
THE GROWTH GAMBLE
Lucent and BT are examples of good practice in harvest venturing.

But our research suggests that harvest venturing units, even when set
up in the right conditions, are frequently unsuccessful.
There is one main reason for them often underperforming: they
frequently attempt to turn spare resources into significant new revenue
streams, a mistake we refer to as the new legs pitfall. The unit may be
set up for harvesting reasons, but is given or develops the additional
objective of delivering new growth opportunities for the parent com-
pany. The new legs pitfall is understandable. It is common to launch a
unit with a brief, such as “have a look in our laboratory and see if you
can use venturing methods to turn some of our technology into com-
mercial ventures. If you spot some new businesses that will give us
additional growth or add to our portfolio, so much the better.” It is also
understandable from the perspective of the managers in the unit. Even
if they are not explicitly asked to look for new growth opportunities,
they are likely to want to. If they find a major new growth platform,
they will be contributing much more to the success of their company.
The problem is that new leg venturing, as we shall outline later, is
not a viable business model. By adding a new leg ambition, managers
in the unit set themselves up for failure and can pollute the effective-
ness of their harvest venturing model. In one company, the desire to
find new growth platforms caused managers in the unit to align them-
selves with projects with “new leg” potential in the hope of building
their career as the businesses grew. Much less attention was given to
finding an exit. When the parent company’s growth ambitions abated
in 2001, the exit window had closed, leaving the unit with no successes
and heavy costs.
Harvest venturing has other pitfalls. Managers in the unit often find
themselves working on poor-quality ideas because the assets they are
trying to exploit are not as promising as originally expected. Existing
businesses or functions hold on to their best assets and only offer less
promising properties to the venture unit. Another problem arises when
managers stay with projects too long in the hope of capturing more
value. Forgetting that their primary objective is to turn existing assets
into cash, they start to treat their portfolio of initiatives as an
213
investment in new ventures. Normally lacking the skills to make the

right decision about exit timing, these managers often overinvest in
their projects and get caught holding the baby when setbacks occur.
These pitfalls make it necessary for harvest venturing teams to stick
closely to their business model (see Box 8.1). A harvest venturing unit
should be cash driven, turning its new ventures into cash as quickly as
possible. The unit should be funded only for its operating budget and
should be measured on cash returns against total funds spent. New
ventures that could become part of the parent company’s portfolio
should be “sold” back to the parent. Assets that the harvesting unit is
working on should be transferred to the unit or there should be a clear
agency agreement between the division that owns these assets and the
harvesting unit.
Some units appeared to avoid these pitfalls, while others learn
through a process of trial and error. Lucent NVG, for example, sold
back to its parent three of its ventures, avoiding the new legs pitfall. BT
Brightstar focused its initial efforts on liberating ideas and technolo-
gies in the R&D organization, but then in 2001 it changed its focus
toward generating a cash return on its ventures. This included refo-
cusing its efforts on technologies that were closer to commercial readi-
ness. Philips’s corporate venturing unit in Amsterdam, which has had
mixed success in spinning off ventures, learnt the importance of getting
control over assets. Rene Savelsberg, the head of the unit, explained
that he made business divisions sign “divorce documents” that ceded
control of assets before he would begin work.
The importance of clarity about the business model is underlined
by the acquisition of Lucent NVG and BT Brightstar by Coller
Capital. Neither unit was primarily cash driven, and in our view
underperformed as a result. Because Coller has no interest in building
new technologies or new businesses for their own sake, we expect its
managers to be more disciplined in executing the harvesting business
model.
214
THE GROWTH GAMBLE
ECOSYSTEM VENTURING
Some companies depend on the vibrancy of a community of connected

businesses for their success. The community may be suppliers, agents,
distributors, franchisees, technology entrepreneurs, or makers of com-
plementary products. Often this community does not need support
from the company other than through normal trading relationships.
However, sometimes the vibrancy of the community can be multiplied
to the company’s benefit by careful ecosystem venturing: by acting as a
venture capital support to entrepreneurs in the community.
Consider the case of Intel Capital. Its development has tracked that
of the microprocessor industry. In the early days, and until 1995, most
Intel Capital investments were made in suppliers, often to guarantee
availability of components. As the industry matured and venture capi-
tal became available for small businesses, Intel turned its attention to
investing in healthy fellow travelers, in order to promote the use of
Intel technology and to act as a window on new technologies.4 This
period coincided with the rapid expansion of the internet, and these
investments helped to reduce the risks that would be associated with
backing the wrong horse.
By 2001, Intel Capital had over $3 billion under management and
was viewed by Intel managers as having contributed significantly to the
success of the company as well as delivering a healthy return on
invested capital. Despite the unsavory market conditions in the follow-
ing three years, Intel Capital continued making investments.
Another example is Johnson and Johnson Development
Corporation (JJDC), which was founded in 1973 as a stand-alone unit
to make strategic investments in the healthcare sector. JJDC makes
minority investments in a broad range of start-up companies, in areas
of technological interest to the parent company Johnson & Johnson.
Very occasionally, these investments lead to acquisitions. But in the
vast majority of cases, the value lies in helping to establish new tech-
nology areas that the company believes in, and in providing the exist-
ing businesses with a window on new technologies. Like Intel, JJDC is
one of the few clear-cut success stories in corporate venturing—it has
215
been around for 30 years, and it delivers returns comparable to a top-

second-tier venture capital firm.
Ecosystem venturing is appropriate when an existing business
depends on the vibrancy of a community of complementary businesses
and the entrepreneurs in the community do not have sufficient support
from existing suppliers of venture capital. This normally occurs
because of the area’s newness. The company may be able to see the
potential, at least in terms of the benefits for existing businesses. But
the venture capital industry may not yet have focused on this area. One
additional condition is that the company doing the ecosystem ventur-
ing needs to be a strong player in its business.
If the community is already well supported, the company with an
ecosystem venturing unit will struggle to gain benefits in its existing
businesses. Moreover, it will have to compete for the attention of
entrepreneurs with existing suppliers of capital. For example, Intel
Capital moved its focus away from suppliers and onto software compa-
nies as the support for and strength of suppliers improved.
If the investing company is not a strong player in its industry, not
only will the returns to its existing business be smaller than those avail-
able to strong players, but also the success rate of the investments is
likely to be lower. A new venture that is building on something the
market leader is doing is likely to be more successful than one that is
building on the activities of an also-ran. As a result, ecosystem ventur-
ing is a strategy used more frequently by industry leaders, such as Intel,
Microsoft, and Johnson & Johnson, than by smaller players.
The major pitfall to ecosystem venturing is the loss of focus tempta-
tion, where the unit managers invest in a wider deal stream and seek
greater autonomy than that which justified the creation of the unit.
This temptation is understandable. Many interesting prospects cross
the manager’s desk and, once a hot investment prospect has been iden-
tified, it is easy to get caught up in “doing the deal” rather than ana-
lyzing the benefits for an existing business. But focus is absolutely
central to making this form of corporate venturing work, because it is
only in a narrow range of technologies that the company has some-
thing to offer over and above the independent venture capital firm. As
216
THE GROWTH GAMBLE
a partner in Siemens’s highly successful Mustang Ventures unit

explained:
The hardest part of my job, the saddest moments, are when we find a
great company to invest in, but we cannot find the alignment to our
business units, so we have to drop it.
To avoid the loss of focus temptation, the venture unit needs to have
clear objectives, in terms of both the sectors in which it is investing and
the relative balance between financial and strategic returns (see Box
8.1). Financial returns are necessary because the venture unit has to
justify its existence to skeptical colleagues, but it is the benefits for
existing businesses that are the unit’s real raison d’être.
One way of reinforcing focus is through performance measures. If
the unit is assessed on its impact on existing businesses, using a ratio
such as “value of benefits for existing businesses divided by invested
capital,” managers are encouraged to focus on the impact benefits and
discouraged from staying with an investment through the second,
third, and fourth rounds of financing.
Another way of promoting focus is to give existing businesses a sig-
nificant level of influence over the unit. The agreement of existing
businesses can be sought before any investment is put forward for
approval, as at Intel and JJDC. Managers from existing businesses can
be appointed to the boards of each new venture, as at Shell’s Internet
Works (an ecosystem venturing unit that closed in 2002). Managers
from current businesses can be used to help with due diligence, as at
JJDC. Managers from the ecosystem venturing unit can be located in
each division, as at Intel Capital.
A form of ecosystem venturing comes under the banner of “a win-
dow on new technology” or “a better understanding of market
developments.” The logic of getting this information by investing in
start-up businesses can be questionable. It may be cheaper and less
risky to get the information by monitoring companies and market
trends. But where the logic is justified, the unit should follow an
ecosystem venturing business model. It should have tight links to the
217
businesses or functions for which it is generating information, and its

performance should be measured in ways that encourage focus. Many
units with a “window on technology” objective fail to stick to the
ecosystem venturing business model and mutate into de facto private
equity or new leg venturing.
INNOVATION VENTURING
The methods of the venture capital industry—rewarding people for

value created, investing in many projects to overcome uncertainty,
stage-gate targets to help with assessing progress, particular care in the
formation of the management team, and so on—can be appropriate
ways of executing some of the functional activities of existing busi-
nesses. Although these methods need to be separated from normal
functional management, they can still be part of the execution of a
broader, more traditional business model. Innovation venturing is the
name we have given to a wide range of uses that managers have found
for venturing-like processes as part of existing functions.5
Shell’s Gamechanger program is one example. Originally estab-
lished in 1996, Gamechanger was set up to increase innovation in the
technical function of Shell’s exploration business. The idea was to take
10% of the technical budget and spend it in a venturing way. Ideas
were identified and screened. Initial finance was provided for promis-
ing ideas, and additional finance was available on a structured basis if
ideas progressed through further screening hurdles. In the following
four years this new approach to innovation was taken up by the tech-
nical functions of other divisions in Shell, and is viewed as having pro-
duced a step change in some areas.
That is not to say that everything has gone well. A number of high-
profile “step-out” projects sponsored by a corporate-level
Gamechanger and operating more like new leg venturing failed to gen-
erate support. But where Gamechanger projects have been in support
of the core businesses, there have been notable successes. By mid-2002,
400 ideas had been screened by Gamechanger, with 32 new
218
THE GROWTH GAMBLE
technologies commercialized and 3 new businesses established.

Gamechanger appears to provide a “purpose” and an “outcome” for
the sort of hunch-based research that scientists demand, but that hard-
pressed business sponsors are loath to underwrite.
A different kind of example comes from Lloyds TSB. The company
was concerned that its strategic planning and budgeting processes, with
a focus on economic value, were discouraging the development of
long-term growth opportunities. Rather than change these processes
and risk undermining a well-developed performance culture, Lloyds
TSB decided to try to solve the problem with a central venturing unit.
This unit, Lloyds TSB Strategic Ventures led by Michael Pearson, is
doing something in a venturing way that might normally be part of the
strategy function. Not surprisingly, it reports to Group Strategy and
Planning. Its role is to identify opportunities within existing divisions
or involving more than one division that are not getting enough atten-
tion, and then, using a venturing methodology, “help” the divisions
develop these opportunities. For some projects the unit provides only
funds and advice. For others the unit takes operating control. Initial
results are promising.
A third example is Nokia’s New Growth Business (NGB) unit,
which was established in 2001 to find and develop new businesses for
Nokia. It carries out large-scale systematic idea generation (for exam-
ple a venture challenge process during 2002 yielded several hundred
new ideas), using a combination of technology-push and market-pull
techniques. Most ideas come from within the company. And it has a
small group of senior executives running the operation, who are free to
invest in and develop any new projects that they believe have long-
term strategic benefits for Nokia. So far, NGB has invested in over 10
ventures, ranging in size from 7 to 60 people. Two ventures are now in
an acceleration phase with their first products coming out.
Innovation venturing is appropriate when an existing function is
underperforming against its potential because there is insufficient
energy directed toward commercially oriented innovation and creativ-
ity. Disruptive technologies, for example, are often best handled
through innovation venturing.6 There must also be some belief that the
219
entrepreneurial energy is latent inside the company and can be fostered

by stimulating “intrapreneurs” or by tapping into external
entrepreneurs.
The belief is that, by providing the right conditions, internal or
external managers with entrepreneurial instincts will take more risks
and invest more energy in developing new technologies or new ways of
working. This involves an acceptance of entrepreneurial behavior,
financial support for entrepreneurial projects, and rewards for
successes.
In some cases the psychological value this delivers to entre-
preneurs—allowing them to pursue their instincts or hobbies and
attributing status to this activity—means that they are prepared to
work, on average, for less total reward. Like farmers, who choose a
career that produces a meager return on capital because of its life-style
benefits, entrepreneurially minded managers may be prepared to work
for less if they are allowed to develop their own projects with a prom-
ise of significant rewards if the project succeeds.
The key pitfall to innovation venturing is to view the unit as a way
of addressing a general concern about lack of entrepreneurial spirit in
the company, rather than addressing a specific opportunity to improve
the effectiveness of a specific function. We call this the culture change
pitfall. During the late 1990s there were countless examples of compa-
nies making this mistake, in industries as diverse as investment bank-
ing, media, consumer goods, and information technology, typically
under the banner of embracing the free-market principles of Silicon
Valley and creating an internal market for ideas, people, and capital.7
To avoid this pitfall, the innovation venturing unit should report to
and be governed by the function it is part of—the R&D function in the
case of Shell’s Gamechanger, the strategy function at Lloyds TSB, and
Nokia Venture Organization. It should be managed by a small, senior-
level team with its own operating budget (see Box 8.1). Of course, the
unit also needs sufficient separation so that the team running it can
make its own investment and development decisions and avoid the
gravitational pull of the existing function. For innovation venturing to
work, it has to balance on a knife edge—if it is too far from the core
220
THE GROWTH GAMBLE
function, it is likely to succumb to the culture change pitfall, and if it

is too close to the core function, it will struggle to carve out a distinc-
tive role.
Obviously, we are not suggesting that managers should abandon
attempts to make cultures more innovative. Rather, we believe that
venture units are not the best tool for such culture change initiatives.
Because it is hard to make venture units successful without close adher-
ence to one of the three viable business models, the lack of financial
success can undermine the culture change effort. Instead of using ven-
ture units to promote culture change, managers are better advised to
focus on training or changes to planning systems or incentives.
PRIVATE EQUITY VENTURING
Some companies set up venture units to work with and compete with
the venture capital industry. The goal is financial: there is no require-
ment that the unit will assist existing businesses or find a new growth
platform to add to the portfolio. The company is doing little more than
diversifying into the private equity industry.
GE Equity is one example of private equity venturing. Set up in
1995, GE Equity grew in five years from a zero base to a fund of $4 bil-
lion invested in 300 companies and a staff of over 200, with offices in
Asia, Europe, and Latin America.8
Another example is Nokia Venture Partners (NVP), which was
established in 1998 as a stand-alone unit in California, and subse-
quently with offices in London, Helsinki, and Israel. NVP makes sig-
nificant minority investments in start-ups in the wireless internet
space, and in all respects it operates as a venture capital firm (e.g., in
terms of fund structures, sequenced investments, and carried interest
for partners). Its success is measured purely in financial returns, rather
than in terms of any particular benefits to Nokia Corporation. Its
biggest success to date was the IPO of Paypal, in 2002. As one of the
partners explained, “We do not do strategic investments [for Nokia]
but the reason we exist is strategic for Nokia.”
221
Private equity venturing is appropriate under rather limited cir-

cumstances. On the surface, it might make sense for a company to get
into the private equity business if it had something significant to con-
tribute. For example, research has been done that suggests that new
ventures supported by corporate investments on average do better than
ones that are not.9 The investment of a major corporation appears to
lend credibility to the new venture. This in turn helps the venture
acquire customers, suppliers, and finance. On this logic, almost any
large company with a good corporate brand could justify getting into
private equity venturing. But this surface logic is flawed.
In order to justify entering the private equity industry by setting up
a private equity venturing unit, a company needs to believe that it has
unique access to a flow of deals. In other words, it has better access to
good deals than normal, independent, private equity companies. When
is this likely? In rather rare circumstances relating to a company’s posi-
tion in a marketplace or a technology. For example, GE Equity started
off with a number of advantages that equipped it for success. GE
Capital, its parent organization, had been doing deals for a number of
years and had access to a promising external deal flow. In addition, GE
Equity invested mainly in projects that could benefit from links with
existing GE divisions, and attracted these divisions with a policy of co-
investing.
The possession of something of value to new businesses, such as a
well-regarded corporate reputation or even access to a market, is insuf-
ficient reason to enter the private equity business, because the “some-
thing of value” can normally be traded either directly with these new
businesses or with an independent private equity company. In other
words, the corporate parent that has a contribution to make can turn
that contribution into value by trading it, without the risks and learn-
ing costs involved in getting into the private equity business.
In addition to unique access, managers also need to believe that the
deal flow they are tapping into is in the early stages of an upswing. All
new venture activity appears to go in cycles. First, some technical
advance or change in law or in consumer tastes creates a new opportu-
nity. Then entrepreneurs begin to exploit the opportunity. Soon the
222
THE GROWTH GAMBLE
new opportunity is heavily overexploited, like a gold seam at the peak

of a gold rush. Failures start to increase and greatly outnumber suc-
cesses as the winners begin to emerge and dominate. Over the next few
years the winners gradually squeeze out the weaker companies and new
entrants fall to a trickle.
To make money in this cycle of boom and bust, investors must
invest early and exit before the shake-out—even if they have invested
in winners. The shake-out period is one of low profitability and high
growth, not usually one that is comfortable for the corporate investor.
Launching in 1996, GE Equity did get into many new markets early
in the boom. In fact, the high percentage of successful units (65%)
may be as much to do with the benign environment as with GE’s
unique access. Nearly everyone did well in private equity during these
years.
The main pitfall in private equity venturing stems from hubris—the
anyone can do this syndrome. Managers become attracted because others
are making a success of something similar. They then enter the busi-
ness, misjudging both the timing and the skills that are needed to cre-
ate a success. Not only do they invest too late in the cycle, often
putting together a second-rate team to compete in a booming sector,
but they multiply their errors by overpaying for poor projects, losing
sight of their unique contribution (if they had one), and holding on to
investments for spurious strategic reasons. The examples are too
numerous and too embarrassing to list.
The best way of avoiding this pitfall is simply to avoid getting into
the private equity business unless the internal logic (the internal con-
tribution and superior access to deals) is so strong that it can be justi-
fied without any evidence of external success stories. There are also
some key elements to the business model that should be borne in mind
(see Box 8.1). First, the unit should be fully separated from the com-
pany, with its own closed-end fund with a short investment period cho-
sen in the light of the current private equity cycle, but of not more than
five years. Second, it should be staffed with seasoned managers from
the private equity industry. Third, the managers should be evaluated
and rewarded as they would be in the private equity world—on the
223
basis of return on investment, and through carried interest in their

portfolio companies.
GE Equity may have been a victim of the anyone can do this syn-
drome. Its early timing was immaculate, but management did not have
a closed-end fund that required early exit and did not reward staff on
carried interest. Moreover, at the height of the boom, managers started
investing in general internet opportunities rather than sticking to the
deal stream where GE had an advantage. Possibly for these reasons GE
Equity failed to cash in its successes before the bust, and by 2003 the
unit had stopped making new investments and was considered inter-
nally to have been of doubtful value.
NEW LEG VENTURING
The fifth type of corporate venturing unit is more associated with an

objective than a particular business model. New leg venturing involves
using venturing techniques to identify and invest in new businesses
that have the potential to become new legs to the portfolio. It is a
remarkably common type: 22% of our examples were categorized at
least in part as new leg venturing.
New leg venturing is common because many companies have core
businesses that are growing only slowly or declining. In their search for
new growth opportunities, managers realize that the pickings adjacent
to their existing businesses are limited. They therefore start searching
more widely and latch on to corporate venturing as a low-cost way of
experimenting and trying out new areas. Moreover, they are encour-
aged in this by some of the most influential academics and manage-
ment gurus.10
Unfortunately, new leg venturing does not appear to work. None of
the units that were wholly or partially categorized as this type was suc-
cessful. Whether the focus was external, using a business model simi-
lar to private equity venturing, or internal, using a business model
similar to innovation venturing, no new legs were created and most
units have been closed down with large losses. Recognizing the
224
THE GROWTH GAMBLE
significance of this finding, we went in search of at least one success

story—and failed. Our success criterion for new leg venturing units
was demanding. We were looking for a unit that had spawned at least
one significant (20% of sales or $1 billion in value) addition to the par-
ent company portfolio. BG’s Corporate Development Division came
close. It screened 24 ideas and launched 11 ventures. Within 18
months one of the ventures—telecoms—seemed about to be a success:
it was thought to have a value of $1 billion. However, the telecoms bust
caused a reevaluation when the company closed the project with a
write-off of £350 million.
The reason for this repeated failure rate is less clear. The logic for
using a corporate venturing approach to explore new avenues for
growth seems sound, but in practice it appears to be flawed. We have
therefore hypothesized five reasons for this type of corporate ventur-
ing not working:
1 Managers only feel the need to use a corporate venturing approach

if the obvious growth paths in adjacent businesses are blocked,
hence the unit is focused on low-probability projects in the first
place.
2 Because the new ventures are developed within a separate unit, they
attract little attention or commitment from the core of the
company.
3 The length of time it takes to develop a successful new division is
longer than most business cycles, causing managers to draw back
from the effort before enough resource has been committed.11
4 If one of the ventures begins to show promise, it starts to compete
for funds with the core. Often the result is to short-change the new
initiative.
5 Early-stage venturing is a tough environment even for professional,
independent venture capital companies. On average, the activity of
these venture capital experts probably earns less than its cost of cap-
ital, even after including the few big successes. Companies that
enter this tough environment without some advantage cannot
expect to beat the odds.
225
As we pointed out in Chapter 2, much of the current literature on

innovation and growth encourages managers to set up venturing
processes and venturing units. The logic is that these units will stimu-
late innovation and that they will provide a better support system for
the large number of new developments that every company should be
sponsoring. Our position in this book has been different. We believe
that managers should not play the numbers game. This is because the
problem of generating new growth businesses is not solved by multi-
plying the number of tries. It is solved when insights about how to do
something better coincide with entrepreneurially minded managers
and a host organization that has the right combination of resources and
sponsorship. Each of these elements happens infrequently. Hence, they
only come together at the same time extremely rarely. Efforts to force
the pace are largely wasted. Rather than trying to stimulate the fre-
quency of each of the ingredients, managers do better to scan the
opportunities waiting for a good one to emerge.
Intel, as we have seen, has had one of the more successful experi-
ences at corporate venturing. This is because its unit was designed
around a venturing model that made sense. McDonald’s has been less
successful. For three or four years it had a unit with a venturing remit,
although not formally a corporate venturing unit. This unit appeared
to be part innovation venturing, helping the core business to be more
innovative, and part new leg venturing, looking for major new busi-
nesses. The fact that the unit did not survive the 2002–2003 refocus
suggests that it was not a huge success.
Instead, in 2003 McDonald’s set up McDonald’s Ventures. In terms
of our venturing models, this is a harvest venturing unit. It contains
five businesses that are the remnants of McDonald’s venturing and
partner brand efforts. The brief to Mats Lederhausen, the unit head, is
to identify whether any of these businesses can be built into significant
new legs for McDonald’s without distracting from the core. For those
that cannot achieve this difficult feat—and we might expect that none
of them can—Mats is expected to find a suitable exit strategy.
226
THE GROWTH GAMBLE
THE ROLE OF CORPORATE VENTURING
This chapter has provided a detailed analysis of five different types of

units that coexist under the corporate venturing banner. We began
with the straightforward observation that most venture units fail, and
in the course of the discussion we highlighted the key pitfalls that lead
to failure in each of the five models. In each of the first four models,
these pitfalls can be avoided and we gave examples of companies that
have been successful. The fifth model, new leg venturing, rarely works.
The single biggest cause of failure is that managers simply fail to define
which model their venture unit is supposed to be following. As a result,
the strategic and/or financial objectives are ambiguous, the structure
and staffing decisions are out of alignment, and the unit’s managers
find themselves being pushed in several directions at once.
One example, from a UK financial services firm, is illustrative of this
problem. Its venture unit was established in 1999 to create new busi-
nesses, and to establish a process for innovative thinking and renewal
across the firm. In the first year, the venture unit manager spent £30 mil-
lion setting the group up, and generating and developing internal ideas.
But without any immediate successes from this effort, and without the
skills to develop really big new projects, he switched focus toward ideas
from outside the company. The manager toured venture capital compa-
nies and investment banks looking for investment opportunities, and he
received several hundred business plans. But again, without a clear focus
on which ones to invest in, the seed funding was not well spent, with only
one significant business taking root. In late 2001 a new chief operating
officer took control of the company and started to focus back on the core
business. The venture unit attempted to rethink itself as an internal con-
sultancy, but it was closed down and the few remaining activities were
folded into the strategic planning group. This venture unit was essen-
tially doomed from the start, because it never figured out its own busi-
ness model—it migrated from new leg venturing to innovation
venturing to ecosystem venturing, all within the space of three years.
Compare this to Nokia Ventures Organization (NVO), which has
succeeded in generating some new businesses and some positive finan-
227
Box 8.1 Key elements of the five venturing models
Venture harvesting
❏ Focus—Generate cash from harvesting spare resources, exclude support to existing busi-
nesses and new leg ideas.
❏ Main pitfall—The new legs pitfall:seeking to develop new growth platforms in addition to
the harvesting remit.
❏ Source of ideas—Mainly internal, but also external VCs and other companies.
❏ How separate—Clearly separate financial unit; separate ownership of resources of “agent”
for primary owner. Report to top management level, often finance. No special governance
required, but a board can be a useful way to involve outsiders.
❏ Skills—Mix of managers: some who understand the resources and some who can “sell” or
“do deals.” Good knowledge of venture capital industry and process of new business cre-
ation. Joint venture skills.
❏ Funding—Operating budgets. Some limited investment funds. Project-by-project funding
for significant projects.
❏ Performance measures and incentives—Cash performance against “allocated”assets.Large
bonuses paid against performance targets. No “carried interest.”
Ecosystem venturing
❏ Focus—Take minority stakes in suppliers, customers, and/or complementors to improve
prospects of existing businesses. Generate value through commercial links with investee
firms.
❏ Main pitfall—The loss of focus temptation: investing too widely and seeking too much
autonomy.
❏ Source of ideas—Mainly external. VCs and direct approaches from candidate ventures.
Ideas linked to existing businesses.
❏ How separate—Separate financial unit reporting to investment board including top man-
agement of existing businesses.Close links to existing businesses through overlapping staff.
Each investment should be sponsored by an existing business.
❏ Skill—Small senior-level team of “investors”: some with strong credibility in the existing
businesses and some with strong credibility in the VC industry.Team must be comfortable
collaborating with existing businesses.
❏ Funding—Operating budgets. Investment funds ring-fenced in operating plan but subject
to project-by-project sanction.
❏ Performance measures and incentives—Significant cash bonus scheme based on impact
on existing businesses and portfolio performance.No “carried interest.”
228
THE GROWTH GAMBLE
Venturing innovation
❏ Focus—Use venturing techniques as a more effective means of performing (part of) an
existing functional activity.Often, but not exclusively, applied to R&D.
❏ Main pitfall—The culture change pitfall: aiming for a broad impact on culture change
rather than a narrow focus on improving part of a function.
❏ Source of ideas—Mainly internal, but also external VCs and other companies.
❏ How separate—Separate financial unit not essential. More of a separate process than a
unit. Report to investment board led by functional director and external (to function)
advisers.
❏ Skills—Small team of “nurturers”: some with strong credibility in the existing businesses
and some with good knowledge of VC industry and process of new business creation. Joint
venture skills.
❏ Funding—Operating budgets.Budget of replaced activity reduced accordingly.Investment
funds ring-fenced in operating plan, subject to stage-gate sanction by investment board.
❏ Performance measures and incentives—Performance benchmarked against rest of func-
tion. Financial interest given to entrepreneurs not to “nurturers.”
Corporate private equity

❏ Focus—Take advantage of a unique deal flow and relevant and non-tradable assets to par-
ticipate directly in the VC/private equity industry.
❏ Main pitfall—The anyone can do it pitfall:believing that it is easy because others are being
successful.
❏ Source of ideas—Mainly external through VC network. Ideas screened against pre-agreed
search specification.
❏ How separate—Clearly separate business and financial unit, located in relevant financial
center(s). Report as other business units. Governance through investment board with a
majority of external directors.
❏ Skills—Primarily VC industry specialists with relevant sector experience: some with in-
depth knowledge and network to tap into the host company’s non-tradable asset.
❏ Funding—Closed-end fund with defined exit date of five years or less.Unit funded through
annual management fee.
❏ Performance measures and incentives—Bonus and share-carry incentives in line with VC
industry norms.
229
cial returns for its parent company. Rather than create a single unit
with multiple or changing goals, NVO has created multiple units, each
with its own highly specific goals and its own dedicated team of
employees. As described earlier, New Growth Businesses is an innova-
tion venturing unit, whose objective is to complement the existing
R&D activities of the businesses. Nokia Venture Partners is a private
equity venturing unit dedicated to providing a financial return by
investing in wireless internet start-ups. In addition, there is the Nokia
Early Stage Technology unit, which invests in promising technologies,
most of which will end up being spun out of the company—a harvest
venturing operation.
The NVO example highlights the key message of this chapter.
There is a role for corporate venturing, but it takes many different
forms. Companies need to choose which form they want and build a
unit to match. Mixed objectives and vague ambitions about building
new legs condemn a venturing unit to muddled thinking and failure.
Success comes from limiting the ambitions of the unit to something
achievable and setting it up with a business model that can deliver.
230
CHAPTER 9
POSITIONING AND
SUPPORTING A NEW BUSINESS
M
uch of this book has been about selection: which new
businesses, if any, should your company invest in? It has
also been about how to organize the search for new busi-
nesses: whether to set up a corporate venturing unit, whether to launch
many or a few experiments, and whether to create a group at the cen-
ter of the organization whose responsibility is the development of new
businesses. We now want to turn to the question of how to make sure
a new business is successful: how to grow the new business into a sig-
nificant division within your company.
Many books on new businesses cover topics ranging from staffing
and resourcing to market entry strategies to the different stages of
growth. In fact, one good book on growing new ventures has chapters
covering selecting, evaluating, and compensating venture manage-
ment; developing the business plan; and organizing the venture.1
Instead of revisiting these topics, we are going to focus on four
corporate-level issues that are critical to success:
1 How far should the new business be integrated with or separated

from the existing businesses?
2 At what level in the organization should the new business report?
3 What support should be provided to the new business by the layers
of management “above” it?
4 How should corporate managers monitor progress, and when
should they pull the plug?
POSITIONING AND SUPPORTING A NEW BUSINESS
We have chosen these four subjects because they are all about the way
the parent company looks after the new business and because they are
all issues where we believe our research has generated some fresh
insights.
INTEGRATE OR SEPARATE?
The received wisdom is that new businesses should be kept separate

from the existing businesses so that they receive dedicated attention
and can develop ways of working, cultures, and strategies that are
appropriate to their customers and competitors.2 If they are integrated
with existing businesses, they are likely to end up with similar strate-
gies and cultures and get less attention and resource whenever the
existing businesses hit a bump.
There is much evidence to support this received wisdom. Charles
O’Reilly and Michael Tushman looked at 35 “breakthrough” product
initiatives.3 Seven of the initiatives were carried out within the existing
functional design. Nine were managed using cross-functional teams
within the existing functional structure. Four were set up as project
teams outside the existing functional structure. Fifteen were set up as
separate business units with their own functional structures. While
90% of the separate business units were successful, none of the cross-
functional teams or project teams was. Only four of the nine “existing
functional structure” initiatives were successful. Moreover, eight of the
initiatives started out with one of the first three structures and then set
up a separate business unit. In seven of the eight cases performance
improved.
O’Reilly and Tushman propose high degrees of separation, what
they call an ambidextrous organization. The businesses are completely
separate for operational purposes but integrated at the next layer up; in
other words, they have the same parent as the existing businesses. In
this way the new businesses are protected from inappropriate influ-
ences from sister companies, but can still draw on some of the
strengths of the parent company.
232
THE GROWTH GAMBLE
Figure 9.1 How to structure new businesses: Burgelman’s matrix
Clayton Christensen also argues for separation rather than integra-

tion when the new business is different from the existing businesses.4
He offers two dimensions of fit: fit with processes and fit with values.
If the new business requires processes and values different from the
existing businesses, it should be set up as a “heavyweight” autonomous
unit. If both processes and values fit the existing businesses, the new
initiative can be part of the mainstream functional structure.
Robert Burgelman, another academic who has studied these issues,
offers a range of structural solutions using some different dimensions
to Christensen. Burgelman argues that the structural solution depends
on the degree of “operational relatedness” and the importance the new
business has for the corporate strategy. He identifies nine different
solutions, showing how each can be linked to these two dimensions
(see Figure 9.1).5
Although each author recognizes that there is a choice of how to
organize and that the choice depends on how different the new busi-
ness is from the existing businesses, the default position is “separate
unless there are good reasons not to.” We agree with this broad rule of
233
Figure 9.2 Degrees of autonomy of new businesses
thumb, especially because managers are often biased against separa-

tion.6 They are often reluctant to make the commitment necessary for
the creation of a new business unit. They believe that the new initia-
tive will be easier to control in the existing structure. They prefer to
keep the new initiative less visible, allowing it to be less accountable for
success or failure. In our database of successes, over 60% operated as
autonomous business units, and most of the rest had some degree of
independence (Figure 9.2).
In practice, the decision is often more complicated than
Christensen and Burgelman’s simple matrices suggest. Other issues
will influence the decision. The availability of management talent, the
need for clear accountability, the impact on the motivation of the man-
agers involved, and the need to be able to change the structure as the
project evolves are all factors that can influence the final organization
structure.
Fortunately, Michael Goold and Andrew Campbell have developed
a framework for designing organization structures that can be used to
address the separate or integrate issue.7 The framework is built on nine
principles of organization design, each of which has been turned into a
234
THE GROWTH GAMBLE
Figure 9.3 Nine tests for organization design
test (see Figure 9.3). For example, the principle around product-
market strategies is that the organization structure should give suffi-
cient management attention to the priorities in the company’s
product-market strategies. The market advantage test therefore has
two parts to it. The first part is about checking that there is an organ-
ization unit dedicated to each market segment that is a priority in the
strategy. The second part is about checking whether any sources of
competitive advantage or important operating tasks require collabora-
tion across these unit boundaries. Where they do, the test asks whether
the organization has processes or mechanisms that ensure sufficient
management attention is focused on making the collaboration work
well (see Box 9.1 overleaf for an explanation of the tests).
This framework of tests makes it possible to do detailed analysis of
the separate or integrate issue. The default position—separate unless
there are good reasons not to—can still be the starting point. In fact,
the market advantage test suggests that a separate organizational unit
is required for any new product-market priority, especially if the new
priority involves some different operational initiatives and sources of
advantage. Egg, the internet bank set up by Prudential, is a classic
example. It was initially established to provide a deposit service to
existing customers, suggesting that it should be a separate production
235
Box 9.1 The nine tests of organization design
Four fit tests

❏ Market advantage test
“Does the design allocate sufficient management attention to the operating priorities
and intended sources of advantage in each product-market area?”
❏ Parenting advantage test
“Does the design provide the layers and mechanisms needed to implement the parent-
ing propositions and strategic initiatives?”
❏ People test
“Does the design accommodate and build on the motivations, strengths, and weak-
nesses of the available people?”
❏ Feasibility test
“Does the design take account of the constraints that may make the proposal
unworkable?”
Five good design tests

❏ Specialist cultures test
“Do any specialist cultures—units with cultures that need to be different from sister
units and the layer above—have sufficient protection from the influence of the domi-
nant culture?”
❏ Difficult links test
“Do all important ‘difficult links’—links between units that will be hard to achieve on a
networking basis—have ‘coordination solutions’ that will ease the difficulty?”
❏ Accountability test
“Does the design make it possible to develop performance measures for each unit that
are appropriate to the unit’s responsibilities, low cost to implement, and motivating for
the managers in the unit?”
❏ Redundant hierarchy test
“Does the design have any levels in the hierarchy or any responsibilities retained by
higher levels that are not justified by a knowledge and competence advantage?”
❏ Flexibility test
“Will the design help the development of new strategies and be flexible enough to
adapt to future changes?”
236
THE GROWTH GAMBLE
unit but draw on the existing salesforces. However, as the ambitions of

Egg expanded to include internet banking and credit cards, it devel-
oped a different product-market focus, different operating priorities,
and different sources of advantage. It needed to become a separate
division within the Prudential group.
The value of the nine tests is that they offer a framework. They
provide a checklist to see whether a new business should follow the
default solution of separation or not. Each of the four fit tests can
suggest that a new business should be more rather than less
integrated:
❏ The market advantage test may suggest more integration if the new
business is selling to the same market, using the same product or
technology, dependent on some common operating initiative, such
as an IT platform, or drawing on a shared source of advantage, such
as a shared brand or shared economy of scale.
❏ The parenting advantage test may suggest more integration if the
new business is part of some overarching corporate- or division-
level strategy. For example, if the corporate-level strategy is to pro-
vide a full range of products to a certain customer base, a business
set up to add a missing product to the range would need to be less
separated from the other businesses than one set up to address a dif-
ferent customer base.
❏ The people test may suggest more integration if the business unit
needs to draw on the skills of particular managers or teams that
cannot be fully allocated to the new business.
❏ The feasibility test may suggest more integration if some constraint,
such as regulation or IT systems, makes separation impractical.
Each of the five good design tests can also suggest that a new business
should be more integrated:
❏ The specialist cultures test is about achieving sufficient separation.

But if this test concludes that the cultures and business models are
similar, more integration becomes possible.
237
❏ The difficult links test often suggests more integration. Where links
with other units are difficult to resolve on an arm’s-length basis
(e.g., transfer pricing, shared resources, coordinated strategies),
partial or even full integration of the units may be necessary to get
the optimum outcome.
❏ The redundant hierarchy test may suggest more integration if the
next level above the new business needs to draw on managers in the
new business to achieve division-wide or corporate-wide initiatives.
❏ The accountability test may suggest more integration if the new busi-
ness’s performance is highly dependent on the actions of other units
in the organization. In this situation it may be hard to make the
managers of the new business fully accountable. It may be better to
hold performance reviews and set targets in combination with the
units on which the new business is dependent.
❏ The flexibility test may suggest more integration in situations where
the market or technology across a group of businesses is changing
fast. Firm boundaries between units may need to be more flexible to
accommodate experiments with selling different product combina-
tions to different market segments.
Mars’s move into ice cream in Europe is an interesting example of how

to apply these tests. Mars in Europe was a strong player in confec-
tionery, pet foods, and some other foods such as rice. The confec-
tionery division wanted to enter the ice-cream market using
technology and product ideas from Mars’s US ice-cream business and
the division’s confectionery brands, such as Mars, Snickers, and
Bounty.
The default position—separate unless there are good reasons—
applied. The market advantage test suggested that a new unit would be
necessary for this new product segment. In addition, the specialist cul-
tures test suggested that separation would be needed because ice cream
might need some different strategies. However, some of the fit tests
pointed in the opposite direction. The division was integrating some
functions, such as manufacturing, across Europe and other functions,
such as human resources, across divisions. Major savings were being
238
THE GROWTH GAMBLE
achieved. Hence the idea of setting up a separate business unit with its
own functions ran counter to the division-level strategy (parenting
advantage test). In addition, the confectionery division did not have
and could not easily locate a manager who would be capable of running
a completely separate business focused on the ice-cream market while
at the same time linking closely to the existing Mars organization (peo-
ple test). Mars had found that recruiting managers from outside at a
senior level rarely worked.
The good design tests also suggested more integration. The new
ice-cream business would be competing with Unilever and Nestlé from
a standing start. If it had its own manufacturing, own sales, and own
marketing, it would have a cost disadvantage to its competitors. It
could only be cost competitive if it leveraged the cost structure of the
confectionery business. This created some potential difficult links
between the new business and the functions of its parent organization
(difficult links test). To get the best cost position, the solution was to
integrate the functions so that the manufacturing was done by the con-
fectionery manufacturing function, sales by the confectionery sales-
force, and so on. There were some exceptions, such as a dedicated
salesforce for seaside resorts.
Once the decision to integrate the functions had been made, the
ice-cream business looked more like a new product within the confec-
tionery business than a new business. But should the ice-cream unit be
a small project team within central marketing or should it be set up as
a separate business unit? The conclusion initially was to establish it as
a separate business reporting to the head of confectionery Europe.
This would ensure that it received dedicated attention and did not get
lost in the priority-setting process within marketing (market advantage
test). It also helped create accountability and motivation (accountabil-
ity test). In addition, it made it possible to attract a more senior man-
ager to lead the unit (people test).
A simple analysis of Mars’s ice-cream ambitions in Europe, using
either Christensen or Burgelman’s matrices, would have reached the
conclusion that some middle-ground solution—neither complete sep-
aration nor complete integration—would be appropriate. But it is only
239
by using the nine tests of organization design that managers can be

confident that they have thought through all the issues and can test
fine-grained alternatives to see which is likely to work best.
REPORTING LEVEL
The choice about who a new business should report to is a subset of

the organization structure choice and hence can also be analyzed by
using the nine tests. In the Mars ice-cream situation, the choice ini-
tially was to have the new business report in at the European level
rather than the marketing level. Later, Mars created a new businesses
division within confectionery Europe. Ice cream, along with three
other new businesses, formed part of this division.
The choice of reporting level is between low down, meaning
reporting into some function or smaller unit, or high up, meaning
reporting into the chief executive. In our database of successes, 63%
reported to the CEO and 24% to a major division. None reported to a
new businesses division or new ventures group (Figure 9.4 overleaf).
The reason for devoting a separate section to the question of
reporting levels is because, as in Mars, this is frequently a tricky deci-
sion and because we have some new insights to share.
The advantages of reporting high up—to the chief executive—are
substantial:
❏ It ensures that significant attention is focused on the opportunity

(market advantage test).
❏ It ensures that the wisdom of the organization is applied to helping
this new business, because the CEO is the person most able to call
on his or her colleagues to help (parenting advantage test)
❏ It makes it easier to recruit high-quality managers (people test).
❏ It makes it easier to provide protection from intrusive functions or
divisions (specialist cultures test).
❏ It can help with difficult collaboration issues between the new busi-
ness and other businesses in the portfolio, because the chief
240
THE GROWTH GAMBLE
Figure 9.4 Reporting relationships
executive is likely to learn about them and has the power to step in
and solve them (difficult links test).
❏ It ensures that there are no unnecessary layers of management
between the unit and the chief executive (redundant hierarchy test).
❏ It provides motivating accountability (accountability test).
The main problems with high-up reporting are the following:
❏ The chief executive is usually already overburdened with reporting

relationships. As a result, it can be hard for him or her to give suf-
ficient time to the new business (feasibility test).
❏ The arrangement is often rather inflexible, because managers who
have reported to the CEO are reluctant to change to a reporting
relationship lower down in the organization at a later date (flexibil-
ity test).
As a result, although the default position is often high-up reporting, a

lower-down reporting relationship can be a better choice.
To help with the choice, we have developed some thoughts that
guide managers to the right compromise between high up and low
241
down. These thoughts are additional to rather than a replacement for

the nine tests.
WHERE IS THE NEW BUSINESS IMPORTANT TO STRATEGY?
First, the new business should report into a layer in the organization
for which the new business is an important part of the strategy at that
layer. For example, the ice-cream initiative at Mars Europe should not
report into Mars Inc. in the USA. It was not a significant part of Mars
Inc.’s corporate strategy. It was a significant part of the strategy of the
confectionery division in Europe, hence this was a good level at which
to report. The European ice-cream initiative might have been a signif-
icant part of the strategy of the broader business in Europe (confec-
tionery, petfoods, and other foods). If it had been, it could have
reported in at a level above confectionery. The initiative might also
have been a major strategy for Mars’s ice-cream business, Dove, in the
US. If it was, then the unit could have reported to Dove.
There is no right or wrong level with regard to the strategy ques-
tion. The rule is purely pragmatic. At which level in the organization
is the new business a major part of the strategy? This is the level to
which the unit should report. If it is important to more than one level,
there is a choice. The Egg initiative at Prudential was important both
to the UK division and to the corporate level. Hence this rule would
not have pointed to a final choice for Egg.
IS THERE A MANAGER WITH THE RIGHT KNOWLEDGE

AND SKILLS?
Second, the new business should report to a manager with the know-
ledge, skills, and time to be a good parent to the business. This is a
restatement of the redundant hierarchy test. The knowledge and skills
needed depend on the industry that the new business is in, the skills of
the management team running the new business, and the nature of the
links between the new business and the existing businesses. The “par-
ent” manager will need to understand the industry in order to be able
242
THE GROWTH GAMBLE
to give good counsel to managers. Understanding the industry will also

help him or her protect the business from inappropriate influences
from corporate functions or other divisions. He or she will need to
have, or have access to, skills that the management team of the new
business is weak in. This might be IT skills or negotiating with regu-
lator skills or financial control skills. Finally, the parent manager will
need to be able to smooth the links with other parts of the
organization.
In the Mars ice-cream example, there was no ideal parent manager
for the unit. The head of Europe did not have particular knowledge of
the ice-cream industry and, although he had most of the other skills
needed, he had limited time available.
HOW MANY LAYERS BETWEEN INVESTOR AND

ENTREPRENEUR?
Third, there should be the minimum of layers between the manager

allocating the money and the manager leading the new business:
between the investor and the entrepreneur. The reason for this is that
the manager allocating the money is the manager who ultimately
makes the decision about whether to support this new business or not.
In addition, as the business develops, this manager makes important
decisions about how much money to allocate to the new venture. This
manager or committee (in some companies the tough allocation deci-
sions are made by a committee) needs to be as close to the new busi-
ness as possible, so that he or she (or they) has as much understanding
of the issues as is needed to make wise allocation decisions.
One of the advantages that the venture capital companies have over
corporate parents with regard to supporting new businesses is that the
venture capitalist has no layers of management. Each partner in the
venture capitalist firm will have his or her own portfolio of investments
and will make the allocation decisions. The decisions are subject to
approval, but the head of the venture is dealing with a principal of the
firm not an agent. Moreover, because the partners in the firm do not
have many operating responsibilities, they are able to spend as much
243
time as necessary getting close to their investments and understanding

the issues. Corporate managers need to try to mimic this way of work-
ing as far as possible.
The disadvantage of having intermediate layers of management is
significant. One problem is that the intermediate layer becomes risk
averse. Since the intermediate layer is neither leading the new business
nor taking decisions about how much support to give the new business,
the role has few upsides and significant downsides. If the new business
does not work out, some of the blame will fall on the intermediate
manager. If it does work out, most of the credit will go to those mak-
ing the tough decisions. This reality normally makes the intermediate
manager risk averse.
An additional problem is that the intermediate manager inevitably
screens information from the layer above. Sometimes this is inten-
tional, but more often it is merely an inevitable consequence of having
the extra layer. Without the extra layer, there is direct dialogue
between the manager in charge of the new business and the manager
making the investment decisions. This direct dialogue inevitably
results in more information flow. Again and again in our research we
encountered problems where new businesses were reporting to an
intermediate layer.
This does not mean that all new businesses should report into the
corporate board. What it does mean is that they should report into a
layer in the organization that has the discretion to allocate money to
this business without being second-guessed by layers above. For small
projects, this may be low down in the structure.
This guiding thought appears to be in contrast to recent work done
by Burgelman.8 Burgelman argues that intermediate layers play a vital
role of communication. They help the managers running new busi-
nesses understand the corporate-level strategy and they help
corporate-level managers decide whether to extend the corporate
strategy to incorporate some successful new initiative.
In our view, these apparently divergent positions can be reconciled.
Where the intermediate manager has discretion to allocate money to
the new venture, it is possible for Burgelman’s and our objectives to be
244
THE GROWTH GAMBLE
achieved. If, however, the intermediate manager is a “middleman”

without the power to approve budgets, then we have observed that the
damage done through risk aversion and information screening out-
weighs the benefits of having an intermediary helping the communica-
tion flow with corporate levels.
DO REPORTING RELATIONSHIPS ALLOW OBJECTIVITY?
Fourth, the reporting relationships should provide some process of

objectivity or governance in order to offset the risk of having the man-
ager allocating the money becoming its champion or major sponsor. As
one chief executive explained to us:
The greatest mistake I have made was getting too close to a manager
who was in love with his business. He got me to fall in love with it too.
The cost to the company was many millions of write-offs.

When the manager with discretion to allocate money to the busi-
ness is not the corporate chief executive, there are normally plenty of
checks and balances. The layer or layers above will provide sufficient
challenge and objectivity to spot when the sponsoring manager has
“fallen in love” with the new business.
When the new business reports to the chief executive, however, the
risks increase. In our sample of major successes, while 63% reported to
and were encouraged by the corporate chief executive, many of the
most expensive new business failures are ones that report to and are
protected by the chief executive.
Nevertheless, the problem does not only exist when new businesses
report to the chief executive. It can also occur when a new business
reports to a powerful member of the top team. If this person is firmly
committed to the new venture, it can be hard for the other senior exec-
utives to act as an objective safety net.
A solution is not always possible. Without a balance of power in the
executive team, strong-willed individuals can ignore the counsel of col-
leagues and overrule objective guidance. Hence when new businesses
245
report to managers on the corporate executive team, especially the chief

executive, the chief finance officer should form a governance board for
this new venture consisting of some executives and some non-
executives. The role of this board is to provide the extra objectivity that
can be so necessary when the new business starts to underperform.
SUPPORT FROM LAYERS ABOVE
New businesses can be new ventures that need a great deal of support,
or acquisitions of established companies that need very little support,
or something in between. It is important, therefore, that explicit
analysis is carried out to define the support required. We call this
Parenting Opportunity Analysis (see Box 9.2 overleaf): it is an analy-
sis of the opportunities that exist for parent managers to help the new
business.
Our approach to providing support to new businesses is highly tai-
lored: we believe that parent managers need to think about each new
business individually and decide both what support is needed and,
importantly, what support is not needed. In our view, one of the main
reasons for new business failures comes from giving new businesses too
much support rather than too little. We are not referring here to
money, but to advice and guidance. Parent managers and functional
experts are either so eager to help the new business or so determined
to control it that they burden it with too much guidance, interference,
and support. Finding the right balance is critical.
When Unilever bought Calvin Klein, a luxury perfume and cos-
metics business, senior managers were concerned that there would be
too much influence on this new business from the Unilever central
functions and the Unilever culture. Unilever businesses were mainly
mass-market consumer products, in contrast to Calvin Klein’s up-
market products. As a result, the culture and ways of working at Calvin
Klein were rather different. In order to provide some protection to
Calvin Klein, Unilever decided to create a gatekeeper. For the first
year all contact between Unilever managers and Calvin Klein
246
THE GROWTH GAMBLE
Box 9.2 Parenting Opportunity Analysis
1 List the major tasks that managers in the business unit will need to complete in the
next period. For a new venture, the period may be six months or a year. For a more
established business the period should be two to three years.The tasks will be items
such as:
❏ get regulatory approval for the new technology
❏ put in a new integrated IT system
❏ enter the German market and win 5% market share
❏ upgrade skills in the supply chain function
❏ reduce the cost of manufacture for product X
2 Examine each task (a typical business unit will have five to ten major tasks) and
assess whether the management team of the business is likely to do the task excel-
lently, well, averagely, or poorly. For tasks scored averagely or poorly, consider what
support could be given to the management team to help it complete the task more
effectively.
3 List other areas of support that managers in the parent layers consider the new
business will need. Start by listing the thoughts of the line manager to whom the
unit reports. Add the thoughts of higher levels of line management. Add to this list
other areas of support suggested by corporate functions,such as finance,marketing,
planning, research, and so on.
4 Review the full list of areas of support and decide which the company is capable of
providing.Then consider this subset and identify the top three to five areas of sup-
port in terms of their likely impact on the performance of the new business. The
objective is to focus the support on those few areas that will really make a differ-
ence.Finally, check that the line manager and supporting functions have the capac-
ity to provide the support needed, and make adjustments either to capacity or to
support plans (or even reconsider investing in the new business at all) to ensure that
the support plan is reasonable.
247
managers had to be channeled through one individual. This is a

process solution to the issue of what support to provide.
In contrast to our view about tailoring support for new businesses,
many authors advise companies to set up a new businesses support
team or new ventures incubator to help new ventures find their feet.
We are cautious about advising companies to do this, because we
observe that these systems do not appear to succeed. Successful new
ventures appear to come about in a much more serendipitous way.
They do not seem to be susceptible to a systematic development
process. Hence when companies set up a structured support group for
new ventures, they normally encourage managers to invest in too many
initiatives with low probabilities of success. If it is possible to maintain
discipline about the number and quality of new ventures, a system of
structured support can be useful.
IBM’s Emerging Business Opportunity program, already men-
tioned in Chapter 7, identified four areas of support that would be
given to each EBO—“the extra support and attention needed by
young, unformed ventures”:
❏ Leadership: traditionally IBM had assigned younger managers to

lead new ventures, but they proved not to have the organizational
clout needed to work across IBM’s complex matrix. One of the roles
of the EBO program was to ensure that each EBO was led by an
experienced IBMer. This had implications for skill building, selec-
tion processes, career path processes, and performance evaluation
processes.
❏ Strategy: IBM’s traditional strategy processes presumed reasonably
stable strategies for each business. With EBOs the strategy might
evolve three or four times in a year. Hence the management teams
were given extra support on strategy development.
❏ Resources: the traditional funding process did not secure funds for
new ventures. The venture leaders had to continuously scramble
and “bootlace” resources. EBOs were given secure funding pro-
grams and support in finding the people they needed.
❏ Tracking and monitoring: IBM’s traditional performance metrics
248
THE GROWTH GAMBLE
were often unsuitable for new ventures. Hence a separate system of

milestone reporting was developed, and finance also ring-fenced the
revenues and costs associated with each venture so that the financial
picture was clear.
IBM’s program is a good example. Even so, we criticize it for being

insufficiently tailored to the particular needs of each EBO. In practice,
no doubt, the support actually given to each EBO was adjusted based
on that EBO’s needs. However, the menu of support areas was devel-
oped top down rather than bottom up and has become institutionalized
rather than driven by the particular needs of the EBOs. It is this pre-
sumption that all new ventures need the same kind of support that is
dangerous.
One particular issue of support worth special mention involves per-
suading existing businesses to offer help to a new business. Existing
businesses may have specific skills—like product design, customiza-
tion, low-cost production, or customer relationship building—that
could strengthen the new business. Similarly, unique assets and
resources, like highly skilled IT professionals or a motivated dealer
network, may be owned by one of the other businesses. Making such
skills and assets available to a new business is generally very difficult.
Even when a new business is linked to an existing business by a com-
mon reporting relationship or some shared functions, the natural
instinct of each unit is to focus on meeting its own objectives. Any sup-
port provided to the new business can be seen as a distraction.
For the new business unit, the ability and willingness to seek and
accept help are not automatic either. When a new unit is being built up
from scratch, it often lacks some of the disciplines and experts that
understand what is needed. In the case of acquisitions, there may be a
mixture of arrogance and wounded pride, which discourages managers
from seeking help.
Corporate managers have a key role to play to overcome these bar-
riers. While we do not have any foolproof solutions, the following
actions can help:
249
❏ Create some shared objectives for existing businesses, which include

building the new business.
❏ Make sure that the progress of the new business is frequently com-
municated to all the managers that need to be supportive.
❏ Structure incentives such that managers in existing businesses are
rewarded for the success of the new business.
❏ Ensure that managers’ personal objectives include transferring skills
and sharing assets with the new business.
❏ Set up councils or networks of experts to foster the transfer of
knowledge and skills to the new business.
In other words, do not leave this crucial source of support to chance. If

corporate managers do not actively promote and police the transfer of
skills and resources, it will only happen in a patchy way.
MONITORING PROGRESS
Acquisitions of established companies can be assessed by the parent

company’s normal monitoring processes. Some additional monitoring
may be needed to ensure that the cost savings and other synergy ben-
efits that justified the acquisition are fully realized, but no significant
change will be required to budgeting and planning processes.
Monitoring new ventures, however, does pose a problem. Not only
are these businesses difficult to monitor against plan or against budget,
but when performance is unexpectedly poor, there is always a question
about whether to pull the plug.
The received wisdom is that new ventures should be monitored
against milestones, rather than plans or budgets. One of the few pre-
dictable features of a new venture is that in its early years, it will miss
budget on a number of occasions and it will need to change plans as a
result of new information about competitors, marketplace, or technol-
ogy. As a result, the best way to monitor progress is to agree milestones
that are three to twelve months away and fund the business plan to
achieve the milestone. When the milestone is reached, another is set
250
THE GROWTH GAMBLE
Box 9.3 Confidence Check for new businesses
Traffic Lights confidence

❏ Value advantage
❏ Profit pool potential
❏ Leadership/sponsorship quality
❏ Impact on existing businesses
Execution confidence
❏ Sales confidence
❏ Technology and operations confidence
❏ Partner, supplier, enabler confidence
❏ Support-from-core confidence
❏ Funding and governance confidence
and funded. If a milestone is missed, a new plan is prepared and the

decision about whether to continue to fund the project is revisited.9
We agree broadly with this received wisdom. However, it poses
some difficult managerial issues. How do managers decide what should
be the next milestone? Moreover, when should they decide to stop
funding the project?
To help with these issues, we have developed a tool called the
Confidence Check (see Box 9.3). Its purpose is to help managers assess
their level of confidence that the new venture will succeed. The tool
has two sections: confidence about the judgments in the Traffic Lights,
the judgments that formed the basis for believing that this would be a
success; and confidence about the ability to do what is necessary to
make the project work.
The Confidence Check can be applied at any stage as a new busi-
ness progresses from its beginning to its final stage as a successful divi-
sion. As each milestone is achieved or difficulties are encountered
along the way, corporate managers need a process for confirming their
commitment. The Confidence Check can be at the center of this
process, which should involve the sponsor, the new business team, and
other senior executives such as the chief financial officer, the chief
251
technology officer, and the chief strategist. As a minimum, the process

should be applied as the following milestones are approached:
❏ Approval of a detailed business plan, after the exploratory phase of

a new business idea.
❏ Development of the technology, prototype product, or service is
complete and trial customers are ready to be involved.
❏ Initial scale-up of the offering has been concluded and results from
customer trials are available.
❏ First expansion phase has been finished and further investment is
required to achieve target profitability.
TRAFFIC LIGHTS CONFIDENCE
The Traffic Lights confidence section involves asking four questions

that parallel the four sections of the Traffic Lights: profit pool poten-
tial, value advantage, leadership/sponsorship quality, and impact on
existing businesses. The four questions are:
❏ Are we confident that the profit pool potential is still sufficient to

justify this project?
❏ Are we confident that our value advantage is still sufficient to justify
this project?
❏ Are we confident that we have project leaders and sponsors of suf-
ficient quality to justify this project?
❏ Are we confident that the impact on the core businesses is such that
this project is still justified?
Just as the Traffic Lights were used in Chapter 3 to help managers

decide whether a new business opportunity had a reasonable chance
of success, they are used again throughout the early life of the project
to confirm that it still has a reasonable chance of success. The
analysis carried out before the project started should be revisited and
the consequences of any major changes in judgment considered. The
easiest way for managers to pull the plug on a project is for them to
252
THE GROWTH GAMBLE
decide that one or more of the Traffic Lights is red. This may be
because misjudgments were made when doing the original
assessment or because the world has changed. For example, the
original judgment may have overestimated the skills of the new
business’s leaders. Alternatively, new competitors may have emerged,
costs of supplies may have risen, or customer needs may have
changed.
EXECUTION CONFIDENCE
The second half of the Confidence Check looks at execution issues.

These are the issues for which milestones will be set. Our contribution
here is to define the categories and sub-categories of execution that
need to be considered. The execution confidence section is a compre-
hensive list of the execution factors that should be reviewed. Not all
factors will be relevant to all new ventures. However, once these fac-
tors have been considered, managers can be confident that they have
asked all the important questions.
We have defined five execution categories, each of which has a
number of sub-categories (Figure 9.5). The arrow in the middle of the
Figure 9.5 Execution confidence
253
figure represents the new business. The circle to the right represents
the marketplace the new business is trying to serve.
One execution category concerns that marketplace. The question
we pose is:
SALES CONFIDENCE: Are we confident that we will achieve

sufficient sales at the price we need to make an acceptable return
(assuming, of course, that we can deliver the product or service at
the cost we are planning)?
There are a number of sub-questions that make up the sales confidence

question:
1 Are we confident that we have a significant value proposition, a tar-

get customer group, and appropriate delivery channels? In other
words, do we have a crisp vision for the business?
2 Are we confident that we can identify the target customers and
access their decision process?
3 Are we confident that the target customers will buy our offer once
we have accessed their decision process?
4 Are we confident that the target customers will pay enough to give
us an attractive margin on the cost of goods sold?
5 Are we confident that the target customer group will be large
enough to give us the volumes we need to earn a decent return on
our fixed costs?
6 And finally, are we confident that we know what sales issues to mon-
itor and what stage gates to set?
The first five sub-questions build on each other. A lack of confidence with
any of these is sufficient to create a lack of confidence with the main ques-
tion. The last sub-question, concerning sales issues and stage gates, is a
final reminder that managers need to feel knowledgeable enough about the
business to know what aspect of the sales challenge requires attention next.
A second execution category concerns the operations of the busi-
ness. The question we pose is:
254
THE GROWTH GAMBLE
TECHNOLOGY AND OPERATIONS CONFIDENCE: Are we

confident that we can produce, deliver, and service our offer and
keep abreast of technology at costs that will give us a good margin?
There are a number of sub-questions that make up the technology and

operations confidence question:
1 Are we confident that we can overcome the technical challenges

involved with this offer?
2 Are we confident that we can overcome the legal, environmental,
health and safety, and related challenges involved with this offer?
3 Are we confident that we can produce, deliver, and service this offer
in the volumes planned?
4 Are we confident that we can sell, produce, deliver, and service this
offer at a cost that will provide an attractive return on capital
invested?
5 Are we confident that we can keep abreast of the technology in this
area?
6 Are we confident that we know what operations issues to monitor
and what stage gates to set?
The first three sub-questions cover the technical and operating chal-
lenges. It may be necessary to add additional questions to this list
depending on the particular operating challenges facing the business.
For example, a high-tech company will have many questions about the
technical challenges, such as whether the intellectual property can be
protected and whether standards around the new technology can be
developed. The fourth sub-question is about whether the costs of pro-
viding the offer will, given the likely price and volumes, result in an
attractive return. The fifth sub-question is about the longer term. Will
it be possible to keep up with likely future technology developments?
The last sub-question, concerning operations issues and stage gates, is
a final reminder that managers need to feel knowledgeable enough
about the business to know what aspect of the technology and opera-
tions challenge requires attention next.
255
A third category concerns partners, suppliers, and enablers. This

category covers all organizations external to the company whose sup-
port is needed in the success of this venture. The question we pose is:
PARTNER/SUPPLIER/ENABLER CONFIDENCE: Are we

confident that we can find and retain the support of the partners,
suppliers, and enablers we need?
There are a number of sub-questions that make up the partner/sup-

plier/enabler confidence question:
1 Are we confident that we have or can find partners/suppliers/enablers

with the world-class resources and ability to develop that we need
from them if they are to help us succeed in this market?
2 Are we confident that our partners/suppliers/enablers will remain
committed to working with us despite setbacks or likely other
priorities?
3 Are we confident that our partners/suppliers/enablers will allocate
sufficient numbers of people in support of our venture and that they
will be sufficiently competent and committed?
4 Are we confident that we have or will have good agreements with
our partners/suppliers/enablers?
5 Are we confident that we know what partner/supplier/enabler issues
to monitor and what stage gates to set?
The first four sub-questions cover the main issues that can lead to
problems with partners, suppliers, or enablers. In a specific venture it
will be possible to identify different sub-categories of partners, suppli-
ers, and enablers and tailor the questions more precisely to the issues
connected to the category. The last sub-question, concerning issues to
monitor and stage gates, is a final reminder that managers need to feel
knowledgeable enough about the business to know what aspect of the
partner/supplier/enabler relationships requires attention next.
A fourth category concerns support from existing businesses. The
question we pose is:
256
THE GROWTH GAMBLE
SUPPORT-FROM-CORE CONFIDENCE: Are we confident

that we will get the support we need from existing businesses?
There are a number of sub-questions that make up this support-from-

core confidence question:
1 Are we confident that the existing businesses are motivated to give

us the support we need?
2 Are we confident that the existing businesses will remain commit-
ted despite setbacks or likely other priorities?
3 Are we confident that the existing businesses have the skills and
resources needed to support us?
4 Are we confident that existing businesses will allocate sufficient
numbers of people who are sufficiently competent and committed
to the project?
5 Are we confident we know what support-from-core issues to moni-
tor and what stage gates to set?
The first four sub-questions cover the main issues related to links with
and support from the existing businesses. There are few new ventures
that pass the Traffic Lights that do not depend in some important way
on support from existing businesses. Thinking through the degree and
capability of the available support is important. The last sub-question,
concerning issues to monitor and stage gates, is a final reminder that
managers need to feel knowledgeable enough about the business to
know what aspect of the relationships with existing businesses requires
attention next.
A fifth category concerns funding and governance, the support and
control that new venture will get from the parent company. The ques-
tion we pose is:
FUNDING AND GOVERNANCE CONFIDENCE: Are we

confident that we have or can set up the funding and governance
we need?
257
There are a number of sub-questions that make up this funding and

governance confidence question:
1 Are we confident that the cash-flow requirements and financial out-

comes will be acceptable to our parent/sponsors?
2 Are we confident we have sufficient funds (capital and operational)
to finance the next stage?
3 Are we confident that there is a governance process in place that
will objectively assess the viability of the project, provide funds even
in the face of setbacks or competition from existing businesses, and
stop the project once the probability of success is too low?
4 Are we confident that we have a governance process that will not
impose inappropriate controls, functional policies, or performance
requirements on the business?
5 Are we confident that we know what funding and governance issues
to monitor and what stage gates to set?
The first four sub-questions cover the main issues related to funding
and governance. The parent organization not only needs to be com-
mitted to providing funds, but also needs to know when to restrict the
flow of funds. As we have seen, some of the most costly failures are due
to a lack of control: the parent company becomes as enthusiastic about
the new business as the management team running the project. But
objective control must not lead to inappropriate constraints and per-
formance requirements. The last sub-question, concerning issues to
monitor and stage gates, is a final reminder that managers need to feel
knowledgeable enough about the business to know what aspect of
funding and governance requires attention next.
These five execution categories have been chosen to direct atten-
tion to five of the main causes for new businesses foundering. Research
by Richard Leifer and colleagues was instrumental in helping us think
through these categories. Leifer led a decade-long research project to
track a number of radical innovations. His book, co-authored with five
colleagues, is a fascinating read.10 In it he emphasizes the importance of
managing four areas of risk: market risk (related to our sales confidence
258
THE GROWTH GAMBLE
Box 9.4 The Confidence Check
Are we confident that:

❏ our value advantage is still large enough to justify this project?
❏ the profit pool is still good enough to justify this project?
❏ the leaders/sponsors of the business are good enough for the project?
❏ the impact on existing businesses still justifies this project?
❏ we will achieve sufficient sales at the price we need?
❏ we can develop the technology and deliver our offer at an attractive cost?
❏ we can find and retain the support of the partners/enablers/suppliers we need?
❏ we will get the support we need from existing businesses?
❏ we have or can set up the funding and governance we need?
category); technical risk (related to our technology and operations cat-

egory); resource risk (related to our funding and governance confi-
dence); and organization risk (related to our support-from-core
confidence). Despite the long time frame of Leifer’s research, most of
the innovations were still in development or had been abandoned by
the time the book was written. Hence his lack of focus on a risk related
to our partner/supplier/enabler confidence category may be due to the
fact that most of the innovations did not get to the commercialization
stage of market development and volume production.
When combined with a reassessment of the Traffic Lights, the full
set of questions provides a thorough Confidence Check on a new ven-
ture (Box 9.4).
The check can be displayed graphically and, since checks should be
made at three to six month intervals, the graphs can plot changes in
confidence as well as levels of confidence (Figure 9.6).
The main purpose of the Confidence Check is to help decide what
should be the next stage gate. Normally this will be a target that
advances confidence along those dimensions where confidence is low-
est. If the confidence scores were similar to those in Figure 9.6, the
next stage gate might be something to do with the quality of the
leaders/sponsors and/or the degree of support being given by the exist-
ing businesses. In the example that lies behind Figure 9.6, the problem
259
Figure 9.6 Confidence Check: Illustrative example
that had led to the reductions in both the leader/sponsor and support-
from-core scores was that a senior manager had moved jobs. This
manager had been the main sponsor for the new business, and she had
particularly good relations with an existing business (which she had
previously run) whose help was needed to release property resources
for the new business. Her move meant that she was no longer in a
position to help the new business, and her replacement was less likely
to be able to deliver support from the existing business or to be a good
sponsor to the new venture. The future of the venture was under
threat.
WHEN TO KILL A VENTURE
The decision to pull the plug on a new venture is the hardest and yet
most important decision that managers face. The Confidence Check
helps make the judgment, but does not substitute for it. When one or
more of the confidence scores is at 2 or below and there is no obvi-
ous stage gate that can be set to raise the score, a reassessment of the
project is needed. The Confidence Check does not identify when the
260
THE GROWTH GAMBLE
plug should be pulled, but it does signal times when a project is in

crisis.
The most obvious reason to stop a project is when a number of con-
fidence categories have deteriorated or have remained low despite set-
ting stage gates to improve them. In the case of the project in Figure
9.6, no stage gate was set. Support from the existing business remained
good, but started to deteriorate as this business began to face perform-
ance pressures of its own. The new manager, however, proved to be an
inappropriate sponsor. His reaction to all setbacks was to cut commit-
ment to the new venture and, if the Confidence Check had been used,
a gradual deterioration would have been evident on most of the cate-
gories. Without the stimulus of the Confidence Check, no crisis was
created and the venture did not get closed completely for three years.
Careful use of the Confidence Check should have created a crisis after
six months, which would either have resulted in immediate closure or
some change to the sponsorship of the venture.
Some projects are hard to break into chunk-sized stage gates. When
Motorola started the consortium to create a global satellite communi-
cation service called Iridium in the early 1990s, it had to launch 66
satellites before it would know whether the service was viable. In the
end it turned out that the offering was inherently too expensive com-
pared to cellular networks. It was only of value to a very specialized
user group operating in remote areas of the globe.
In other situations, even in high-tech industries, the Confidence
Check would be valuable. When Philips developed interactive CD
technology starting in the mid-1980s, a Confidence Check process
might have helped it pull the plug on the project five years sooner. The
project foundered because the internet was a superior technology, but
Philips did not close the project until the mid-1990s.
The Confidence Check is not only a tool for cutting back on proj-
ects, it can reinforce confidence in further investment. Philips’s Navteq
project, to develop a navigation service for car users, frequently missed
its milestones. However, the strategic case remained strong.
Substantial investments and management support finally led to success,
creating several billion euros in shareholder value for the company.
261
Careful analysis, similar to the Confidence Check, at critical points in

the development of the venture was key in reaffirming the ultimate
potential and reaching the positive decisions.
KEY MESSAGES
In this chapter we have addressed four of the common issues that man-
agers face when developing a new business. We have focused on new
ventures rather than acquisitions. Our messages have been:
1 Separate the new venture from existing businesses unless the nine
principles of organization design suggest otherwise.
2 Make sure that the new venture reports to a level in the organiza-
tion where:
❏ the new venture is an important part of the strategy;
❏ there are managers who can be good sponsors;
❏ there are the minimum of layers between the new venture and
the “investor”;
❏ there is some objective governance of the venture.
3 Use Parenting Opportunity Analysis to decide what guidance and
support the new venture needs and limit parenting to these ele-
ments.
4 Use the Confidence Check to assess progress, set stage gates, and
engineer crises that lead to reassessments.
262
CHAPTER 10
AN AGE OF REALISM
I
f the messages from this book become common wisdom, we will
have a future that is different from the past.
We will have many companies living comfortably with busi-
ness plans that are low growth. Since the end of the 1960s, when the
first conglomerates began to emerge, such as LTV in the US or Slater
Walker in the UK, Anglo-Saxon economies have been living a delu-
sion. The delusion is that good managers can create growth businesses
even if they are in low-growth markets. Despite the demise of the
1960s heroes, like Jimmy Ling of LTV and Jim Slater of Slater Walker,
there have been enough others who have produce sufficient amounts
of temporary magic to feed a belief that anything is possible.
When we started this project we were asked by managers sponsor-
ing the work to study the magicians of the time: Enron, Vivendi,
Marconi, and others, which were the LTVs and Slater Walkers of the
1990s. Fortunately, the length of our research period helped us see that
these high-growth stars were not exemplars. It helped us become more
confident that endless growth is a delusion.
There are still many purveyors of this delusion. Analysts, managers,
financial journalists, consultants, and academics have all made contri-
butions. Some spread it when they argue that there is no such thing as
a mature business, only mature mindsets. Some reinforce the delusion
when they suggest that low growth is a symptom of a lack of innova-
tion and creativity. Low growth, they argue, could be eradicated if
managers were only more creative and if they would only invest more
in innovation. Some champion technological innovation. Others
champion business model innovation. Both are positioned as a pathway
to a new economy—an economy where profits are easier to earn and
AN AGE OF REALISM
growth is available to all. The delusion fueled the dot-com boom and
it has rocket-powered many management teams to their doom.
Arrayed against this delusion are some wise heads, like Lou Gerstner,
whose criticism of Wall Street we reported in Chapter 4. He resisted the
siren songs tempting him to grow revenue faster by developing into or
acquiring into the “new economy.” Another wise head is Warren Buffett,
the enigmatic leader of Berkshire Hathaway. He has long focused on
quality products and market positions rather than growth. He famously
declared that he could not understand the “new economy” and, to the
huge advantage of his investors, let it and the delusion it rested on pass
him by. Another force for good is the rise of share buybacks. Even Bill
Gates’s Microsoft has joined the stampede to give cash back to share-
holders rather than use it for getting into new businesses.
As these forces for sanity receive intellectual support from books
like ours, we will enter an age of realism: one where managers are com-
fortable managing businesses through long periods of low growth, and
pension funds are delighted to take home the excellent and dependable
returns that these companies are able to deliver. Management teams
will no longer feel they have to have a growth plan, and tools like gap
analysis will become less influential than the Traffic Lights.
In this delusion-free future there will still be plenty of growth com-
panies. There will be as many management teams capable of develop-
ing or acquiring new businesses as there are today. In fact, in the future
there may be more than in the past.
This will not be because of more opportunities for growth into new
businesses, but because managers who have learned the rules of the
game will be better at spotting and developing the opportunities that
fit. Today managers are caught in a round of frenetic activity that inter-
feres with their ability to learn about the best opportunities. In the
future managers will take a more considered, more strategic approach
to new businesses. Rather than squandering money and management
time on too many initiatives, often running out of patience and
resources to give the support that their really promising opportunities
most need, managers will invest only in new businesses that deserve
their attention and hence will give each the attention it deserves.
264
THE GROWTH GAMBLE
In the age of realism, managers will be more patient. They will rec-
ognize that good ideas are rarely good ideas if they do not come to the
table with good managers to lead them. They will spend more time
examining their value advantage, and considering whether they have
sufficient advantage to overcome the unique qualities of their com-
petitors and the costs of learning the new business. They will be more
sensitive to the dangers of distraction, and will avoid investing in new
projects when their existing businesses are likely to absorb most of
their spare resources.
In the age of realism managers will be more creative with their
existing businesses. With less distraction from greener grass, more
energy will go into exploiting the full potential of existing businesses.
The restless, innovative managers who champion new things will not
be banished into business development divisions and corporate ventur-
ing ghettos, they will be embraced in the core or encouraged to seek
their fortunes elsewhere. And if Gary Hamel and his colleagues at
Strategos are successful at doing for innovation what Joseph Juran did
for quality, the effort managers put into innovation will produce dra-
matically better returns. While we do not believe that existing busi-
nesses can be kept growing for ever, few are given the attention they
need to achieve their full potential.
In the age of realism a huge amount of society’s wealth will be saved.
Predictions that progress will stall and innovation will be abandoned
will prove unfounded. Instead, more money will be available for inno-
vation because less will be squandered by overambitious companies.
Just as the market system’s greater efficiency has made more money
available for innovation than the planned economy, the age of realism’s
avoidance of foolish investments will release more money to support
those with good ideas and matching skills.
In today’s world our economy takes two steps back for every three
steps forward. For every Dell or Samsung, there is a maturing com-
pany squandering millions of dollars in failed attempts to enter new
businesses, and at the same time letting slip billions of dollars by
becoming distracted from optimizing their existing businesses. In fact,
we estimate that companies in decline often use up more than half of
265
AN AGE OF REALISM
their market capitalization in unsuccessful attempts to regain their

high-growth youth.
In the age of realism, managers will manage low growth or decline
when it is appropriate, possibly even without the heavy debt structures
of leverage buyouts and private equity deals that seem to be the only
way to get managers to focus on their existing businesses in today’s
growth-addicted world.
The age of realism will still have failures. There will still be projects
like Motorola’s Iridium, where a promising satellite technology was
made uneconomic by the development of cell-based systems. Problems
like Philips’s difficulties in flat-panel displays, where rising Asian com-
petitiveness made it hard to compete from a European base, will still
occur. But they will be less frequent because managers will be more
thoughtful about whether they have sufficient value advantage,
whether the project leaders and sponsors are good enough, and
whether the distraction risks are manageable. There will be less macho
and more measured behavior.
In the age of realism strategy will become a more respected disci-
pline. Currently there are still too many managers, consultants, and
even academics who believe that strategy is the work you do to figure
out how to achieve your objectives. If your objective is to grow at 15%
per annum, strategic planning is about figuring out how to do it.
In the future strategic planning will be about defining both objec-
tives and action plans. The focus will be on plans that have a good
probability of creating value for the organization. Not all organizations
will be driven primarily by shareholder value. Some will have employee
value, societal value, even environmental value as part of their mission.
But whatever value an organization is trying to create, the role of the
strategist will be to help form objectives and action plans that are likely
to succeed. Gone will be the days when companies tolerate strategies
that have a 90% failure rate. Not all strategies will succeed, but the
success rate will improve dramatically.
In the age of realism, the strategic business case will have as much
status in decision making as the financial business case. This will make
optimistic forecasting less necessary and easier to spot. Instead of
266
THE GROWTH GAMBLE
treating planning as a game of “dare”—I dare you to promise high sales

and high margins and I will give you a forfeit if you fail and a prize if
you succeed—managers will agree plans for new activities based on
their strategic logic.
In the future, the strategy function will develop a harder edge. It
will guide management thinking and hold a veto on major decisions.
With the power of tools like the New Businesses Traffic Lights, it will
act as a gatekeeper, only allowing strategies that pass the test of real-
ism. To achieve this status strategy will need a new power base within
the finance function. In fact, the finance function may finally complete
its transformation from an accounting and control activity to becom-
ing an evaluator of strategic plans and a monitor of progress.
In the age of realism the finance function will be the keeper of a
realistic approach to business. Gone will be crude gap analysis, which
has led so many companies into deciding that they need new businesses
to fill the gap between their expected growth rate and their top-
quartile ambitions. Gone also will be simplistic objectives based on
“what the shareholders expect,” derived from the investor relations
activity. Gone will be stretch targets 10% greater than last year’s
budget. In their place will be a realistic analysis of what the existing
businesses can deliver and how much management attention they need.
Added to this will be a sober assessment of the opportunities that exist
to enter new businesses and whether any of these opportunities fit with
the resources the company has to hand. When combined, these two
realistic assessments will help managers define objectives and action
plans for the next period.
The age of realism will be a new enlightenment for business, where
rational thought has more weight than chief executive ambition, and
where managers at many levels have the tools to challenge weak think-
ing. Managers, customers, and shareholders will all benefit. In the
same way that the philosophers of the eighteenth century helped
undermine the dogma of churches and a belief in monarchs, the strate-
gic ideas developed over the last 20 years are helping undermine the
dogma of growth and a belief in corporate magicians. Fortunately, the
transition will be a lot less bloody.
267
APPENDIX A
ADVICE FROM OTHER

AUTHORS
I
n this appendix we summarize the advice that we believe other
authors might give to McDonald’s, taken from a range of excellent
books and articles. We chose McDonald’s as the focus, rather than
Intel or any other company, because we thought that the challenge facing
McDonald’s in the period covered by this book was particularly interest-
ing: a hugely successful core business that had hit a bump or two, an
acknowledged management team, but question marks about the future.
We recognize that the authors we have chosen were not writing with
McDonald’s in mind, hence our summaries are dangerous presumptions.
However, this is the advice that we think a manager from McDonald’s
would have gleaned from a careful study of the authors concerned.
After each summary, we have commented on those aspects of the
advice that we think are good and those with which we take issue. These
comments are necessarily unfair. With one or two exceptions, we have not
given the authors the opportunity to debate the issue with us. However,
we include the comments because we believe they provide readers with
further insights about where our ideas support those of other authors and
where they differ. The exceptions are Gary Hamel and Robert Burgelman.
We have had quite a lot of dialogue with both and they have influenced
our summaries of their advice.
The order is alphabetical by the last name of the first author. We do
not include a section on Clayton Christensen’s work because it is covered
fully in Chapter 6. Other authors whose books we have studied and been
influenced by include Julian Birkinshaw (Inventuring), Don Laurie
(Venture Catalyst), Heidi Mason (Venture Imperative), and Michael Treacy
(Double-Digit Growth). They have not been included because of space
constraints.
ADVICE FROM OTHER AUTHORS
Merhdad Baghai, Stephen Coley, & David White, The Alchemy

of Growth: Practical Insights for Building the Enduring
Enterprise, Orion Business Books, 1999
These authors are from the consulting company McKinsey. They imply
that McDonald’s should maintain a continuous pipeline of new business
initiatives. They propose that companies should think about three sections
of this pipeline: improvements to existing businesses, new enterprises
coming on stream, and future options. They refer to these three parts of
the pipeline as three horizons.
As a company “under siege”—with challenges in all three horizons—
McDonald’s, they suggest, needs a structured program:
1 Step 1 is to earn the right to grow by sorting out the core business and
divesting any distracting or underperforming businesses.
2 Step 2 is to develop the commitment to grow. It may take two years or
more. It may involve management changes. It will involve analysis and
debate. But it needs to be a passion accompanied by setting high tar-
gets and transformational changes to the culture and systems.
3 Step 3 is to explore new opportunities along seven degrees of freedom:
grow share, grow geographically, market existing products to new cus-
tomers, market new products to existing customers, develop new deliv-
ery approaches, consolidate the industry, pursue opportunities outside
the industry.
4 Step 4 is to launch a number of exploratory new businesses in a small
way. The objective is to create a staircase of steps that will build capa-
bilities and expand the opportunity that has been selected. Capabilities
can be competencies, privileged assets, growth-enabling skills, and spe-
cial relationships.
5 Step 5 is to focus on those new businesses where a competitive advan-
tage can be sustained: where the company has built more capabilities
than rivals and has developed a positional advantage.
6 Step 6 is to differentiate the management system for each horizon of
the total pipeline in terms of talent management, budgeting and plan-
ning, and performance management.
270
THE GROWTH GAMBLE
COMMENTS
We find much in The Alchemy of Growth that is compelling. The emphases

on the core business, on capabilities, on how to build capabilities (the
staircase), and on competitive advantage are vital and well addressed.
However, we have concerns about the three-horizon concept, about the
advice to create different management systems, and about the emphasis on
developing a commitment to grow.
The three-horizon concept seems almost tautological: new businesses
take five, ten, even twenty years to reach a significant size. Hence, unless
your company has some new businesses in the pipeline, growth for the
next few years will have to come from existing businesses or acquisitions.
However, the advice that these authors give seems to us to be wrong.
Moreover, their conclusions do not appear to be supported by their data.
They take their three-horizon concept and combine it with the fact
that most new business ideas fail. They then recommend that companies
such as McDonald’s should have a large nursery of emerging new busi-
nesses (horizon 3) in order to have a few fast-growing new divisions (hori-
zon 2) in order to have one significant new core business (horizon 1). This
causes them to recommend that the company build three different man-
agement systems—one for each horizon. The three different portfolios of
businesses (core, growing, and emerging) require different people, per-
formance targets, and planning processes.
We take issue with these recommendations. Only a tiny proportion of
our success stories started out as part of a portfolio of new initiatives. Few
of the new leg venturing nurseries we studied created new legs. Moreover,
we have not found a company that has successfully been able to manage
three different management systems for any length of time. Interestingly,
the authors make the same observation: “Yet as far as we know, no com-
pany, not even those on our list of sustained growers, is using distinct man-
agement systems for each horizon. There are no best-practice cases.”
So what is wrong with the three-horizon concept? If we think about
the growth challenge of some existing businesses, for example a pharma-
ceutical company, there will be three horizons of new product plans.
There will be new products being launched this year, products under
development for launching in the next two or three years, and research on
new products that may not be launched for five or more years. But these
three horizons do not need three management systems. The three
271
horizons work together because they can be managed within the same
management system. This is because there is one business model: running
a pharmaceutical company.
Unfortunately, the same is not true for portfolios of companies. For
McDonald’s, the three-horizon concept is not useful. Like almost every
other large company, McDonald’s will not be able to sustain a different
management system for emerging businesses for long enough for them to
become new legs to the portfolio. In our view, if McDonald’s is to find a
new business, it must find one that will fit with its existing management
system.
Our final concern is the advice to develop a “commitment to grow.”
We believe in a more cautious approach. Commitment should develop
only after an opportunity has passed the Traffic Lights. Commitment to
grow independent of whether there are good opportunities is the prime
cause of many expensive failures.
Zenas Block & Ian MacMillan, Corporate Venturing, Harvard

Business School Press, 1993
Block and MacMillan’s book is the oldest of those selected. The authors had
been studying venturing at Wharton for a number of years and produced a
comprehensive summary of thinking at that time. They published just after
the second wave of corporate venturing initiatives and their work shares a
sense of optimism for the possibilities and frustration at the failures.
McDonald’s, they suggest, will need to make a significant cultural
change in order to become more entrepreneurial and create new busi-
nesses. The change will take five years or more and is likely to require
alterations in top management. Three elements are critical:
❏ Leaders who can articulate a unifying vision and strategy.

❏ A culture that supports initiative and entrepreneurship.
❏ Skills in managing a venturing effort and in developing individual
ventures.
The venturing effort will require senior managers at McDonald’s to
❏ decide that venturing is going to be an integral part of the corporate-

level strategy;
272
THE GROWTH GAMBLE
❏ choose between a venturing strategy based on extending existing capa-

bilities into adjacent areas or on developing technology in new fields;
❏ determine the size of the effort in terms of number of ventures and
investment commitment;
❏ lead the development of ideas for new ventures;
❏ develop and disseminate a set of venture selection criteria and explain
decisions to managers involved;
❏ set up each new venture so that it draws on the organization’s strengths
and avoids its weaknesses;
❏ prepare for disappointments by using milestones to reset strategies,
“shooting the wounded” early, distinguishing between bad luck and
bad decisions, and setting up a process for learning from success and
failure.
Half-hearted efforts involving a few new ventures or a separate corporate

venturing division are unlikely to succeed. McDonald’s needs to “sign up
for the course” and make sure that all of its divisions have a venturing
strategy.
COMMENTS
Block and MacMillan’s book is one of the few that covers the tasks of both
senior managers and venture-level managers. Much of what they say is
based on deep research and, rightly, remains unchallenged.
If a company does decide to develop a venturing program, Block and
MacMillan provide plenty of excellent advice. However, our views differ
in some fundamental ways. Our main criticism is that they do not say
which companies should set up venturing programs and which should not.
In fact, they argue that all companies should. We do not think that a ven-
turing program, of the form suggested by Block and MacMillan, is appro-
priate for most companies. In fact, we do not believe that corporate
venturing programs are a successful way of developing new businesses.
Nor do we believe that most companies are capable of building the skills
that Block and MacMillan argue they will need to succeed. We are also
strongly against advising a company like McDonald’s, which has focused
for 50 years on one business, to launch a cultural change aimed at making
every division develop a venturing program.
273
At a more technical level, Block and MacMillan have observed that

companies are more successful when top managers are committed, pro-
viding a clear strategy for the effort based on a deep understanding of the
company’s capabilities and weaknesses. Possibly this is where our views
coincide. We also encourage managers to do top-down thinking. They
need to identify areas where the company’s capabilities will give it an
advantage. We then suggest that top managers pick one or two of the most
promising investments. Where we part from Block and MacMillan is their
suggestion that managers will need to launch many rather than few new
investments and that they need to build a generic process and support sys-
tems to nurture these ventures.
In our view, the number of new ventures, at any one time and at any
one level in the hierarchy, will be so few that it is not appropriate to build
a new businesses development capability. Instead, managers should focus
on what is required for each of the unique investments that are being sup-
ported. We think in terms of a few investments each treated in a unique
way, while Block and MacMillan imply a flow of investments processed
through an organizational capability.
Even on this point, our differences are more about degree of effort. We
support the creation of a small team at the center to help managers screen
ideas, set up new ventures, and give support when needed. An analogy
might be the M&A team that exists at the corporate level of most compa-
nies, whose role is to help screen prospects and assist with the deal. What
we are not comfortable recommending to a company like McDonald’s is
that managers set up a venturing program before they have decided if
there are any ideas worth investing in. It is better, we argue, for managers
to see if there are any ideas that pass the Traffic Lights. If there are two or
three in each of five geographic divisions, then it may make sense to set up
a small team at the center to offer support to the division managers
responsible for these investments. In this way the capabilities are built to
support the immediate task rather than to develop a generic organizational
competence.
274
THE GROWTH GAMBLE
Robert Burgelman, Strategy Is Destiny, Free Press, 2002
Robert Burgelman is a professor at Stanford Business School who has

published widely on growth and new businesses. Strategy Is Destiny sum-
marizes his thinking using detailed analysis of Intel’s attempts to find new
sources of growth in the 1980s and 1990s. The implications for
McDonald’s are as follows:
1 McDonald’s top management must view the capability to develop new

businesses not as “insurance” against the core business becoming insuf-
ficient to sustain profitable growth, but as an integral part of the long-
term strategic leadership capability of the company. This will avoid the
predictable fluctuations in support for corporate entrepreneurship that
lead many companies to go through cycles of “now we need it; now we
don’t.” The consequence of such cycles is that each time a new one is
initiated, the company starts from scratch.
2 McDonald’s top management must understand that new business devel-
opment requires a different type of strategic leadership discipline than
the strategic leadership discipline associated with the core business.
Both disciplines must be pursued simultaneously, for the long term.
3 The key activities necessary to develop a new business involve different
levels of management. At the operational level, the activities involve
linking technology with customer needs, championing new ideas, and
“strategic forcing.” At the middle/senior level, the activities are “strate-
gic building,” championing organizational solutions, and delineating
the boundaries of the new business in the new industry. At the top level,
the activities are structuring the organization and engaging in “strategic
recognition” and retroactive rationalization. These top management
activities are not “reactive.” Rather, they require the intelligence to sus-
pend the rules for some time and the tolerance to embrace ambiguity.
Only when major uncertainties (technical and commercial) are reason-
ably resolved and support is building among a number of (though not
all) senior executives should senior managers commit the corporation to
a new business. Learning to perform these activities on an ongoing basis
is part of developing the corporate entrepreneurship discipline.
4 Companies are successful in new businesses when their strengths and
weaknesses match well (i.e., better than competitors) the critical success
factors in the new business.
275
5 The first step McDonald’s should take is to search for new businesses
that have “appropriate” critical success factors. This can be done by
managers at all levels, but is only likely to be successful if the screening
process includes an accurate understanding of both the strengths and
weaknesses of the company and the critical success factors of the new
business. Early attempts to enter new businesses may, therefore, be
highly educational in terms of helping managers understand their
strengths and weaknesses and their ability to judge critical success fac-
tors. In other words, managers should not be dogmatic about what the
existing corporate strategy says about their strengths and weaknesses.
Managers should be prepared to experiment a little.
6 If it is necessary to look beyond businesses with “appropriate” critical
success factors, the company is taking on a tough challenge that, if suc-
cessful, is likely to be as much evolutionary as planned. Progress
toward the goal will involve:
❏ leading some top-down strategy work to help develop a new corpo-
rate strategy and initiate the building of some new capabilities (but
recognizing that top-down initiatives do not have a superior track
record);
❏ being sensitive to and maintaining links to the autonomous actions
of middle and junior managers who want to grasp opportunities and
develop capabilities (but recognizing that these initiatives can often
draw the company into major problems);
❏ balancing the dissonance that is likely to arise between efforts to
encourage new initiatives and efforts to drive development of the
core (particularly hard when the core is performing well and there
is little spare resource);
❏ redefining the strategy. This ill-understood part of the new business
development process links the new business to the corporate strat-
egy, thereby amending it. Such amendments are intrinsically diffi-
cult. They involve the combined activities of middle/senior
managers and top managers, as mentioned above.
7 Use a new business development group or internal venturing unit as a
transition vehicle for helping new ventures find their “right” degree of
autonomy and linkage with the core. This requires viewing new busi-
ness development as a discovery process and having available an array
of organizational options, from locating the activity within an existing
business to setting it up as an independent unit with external share-
276
THE GROWTH GAMBLE
holders (see Burgelman’s Designs for Corporate Entrepreneurship frame-

work). As a new business develops and the nature of its interdepend-
ence with the core becomes clearer, the appropriate type of
organizational design may change. Sometimes the right solution may
be to spin off the new business completely. Managerial measurement
and reward systems must be flexible enough to continue to reflect the
changing interdependence and help maintain the necessary collabora-
tion between the new business and parts of the core business.
Ironically, the greatest resistance to this part of new business develop-
ment often comes from the human resources function. The legal func-
tion may also raise concerns about some of the contractual
arrangements that may be involved. The greatest impediment to evolv-
ing the right organization can be the impatience of energetic top man-
agers, who want to see results right away.
8 Use corporate venture capital to stimulate the exploration of
“autonomous” opportunities that are occurring in the external envi-
ronment. But keep in mind that external ventures will raise integration
issues just like internal ventures, which again indicates the important
role of the processes of “strategic context” determination.
9 Work on spotting, developing, and deploying middle and senior-level
managers capable of nurturing new businesses and envisioning a role
for these new businesses in a new corporate strategy. This should be
part of the mandate of the executive development function in all estab-
lished companies.
COMMENTS
Robert Burgelman is one of the few authors who does not focus on the
numbers game. He also avoids the rhetoric of optimism that pervades
most of the literature. However, he does suggest that all companies must
learn the “disciplines of developing new businesses,” which is a rather dif-
ferent position from ours. He also encourages more autonomous initia-
tives (skunk works and experimental ventures) rather than fewer.
Burgelman’s work on autonomous initiatives is a huge contribution to
the field. We agree with much of what he has found. But we take issue with
his implication that managers should encourage more experiments and
initiatives rather than fewer. In a debate in the European Business Forum,
277
Burgelman supports his argument by suggesting that new ventures are

part of “the irrepressible human spirit.” We agree, which is why managing
in this environment is difficult. Going with the flow of the human spirit is
not necessarily the most value-creating strategy.
Burgelman points out that the disciplines of continuous development
are different from those of normal strategic planning. He therefore argues
that companies need to learn these additional disciplines, and that they
need to learn how to do both these new disciplines and normal strategic
planning simultaneously.
We are uncomfortable with this recommendation. There are compa-
nies whose business is developing new businesses: Virgin, 3M, maybe
Canon, venture capitalists. However, we believe that there are no compa-
nies that have the ability simultaneously to run a core business and con-
tinuously develop new businesses. Hence Burgelman notes the stop–go
behavior of so many companies. This signals to us that he is recommend-
ing a solution that he has not observed in practice. Intel’s efforts, his main
case study, have not been a success.
In our opinion Burgelman is trapped by the seeming tautology that
companies must develop new businesses “for the long term” and that
bottom-up initiatives are irrepressible. Since existing businesses have
finite lives, he argues that a company must develop the skills to harness
these bottom-up initiatives if it is to survive. Hence, he makes the seem-
ingly obvious recommendation that companies need to develop the
appropriate disciplines.
Our view is that all companies have finite lives. Long-term survival,
while a natural desire, is not necessarily a rational or value-creating goal.
Developing the ability to create new businesses is not an imperative and
may be inappropriate for many companies at certain points in their life
cycle. Supporting bottom-up initiatives is often value destroying.
McDonald’s may be able to develop another significant business, but it is
unlikely that it will create value from launching a major effort to develop
the skills that Burgelman suggests.
Burgelman believes that most companies cannot solve their growth
challenge by top-down analysis. In fact, he argues that initiatives backed
by the CEO are dangerous because they garner too much commitment.
He believes in a much more organic process in which autonomous actions
lower down in the company discover the way forward. The task of top
management is to stimulate an organic bottom-up process, suspend nor-
278
THE GROWTH GAMBLE
mal judgment, and prepare to integrate successes into a new corporate

strategy.
Our view is that the problems Burgelman observes in top-down think-
ing are problems that are easier to cure than he believes. We see them as
a failure of top-down wisdom rather than a failure to implement all the
disciplines that he suggests. In our shadowing of managers, we observed
many occasions where the thinking of the managers and their most senior
bosses was strategically flawed in quite elementary ways. Top-down think-
ing is often poor—as we believe it was at Intel before Andy Grove—but it
is relatively easy to change (with models like the Traffic Lights and the
Confidence Check). Rather than encourage managers to invest in nurtur-
ing a new set of disciplines that require new behaviors at many levels in
the company, we believe that encouraging managers to improve their top-
down thinking has a better chance of success. In fact, without an improve-
ment in top-down processes, Burgelman’s advice, like that of Hamel, will
lead companies into supporting too many bottom-up initiatives (as Intel
has during much of its past).
Burgelman’s use of “critical success factor” fit as a screening device for
new business ideas is an advance on the usual “core competence” or “adja-
cencies” thinking. It recognizes the learning costs of getting into busi-
nesses with different critical success factors, a point that we make much of
in the Traffic Lights.
Richard Foster & Sarah Kaplan, Creative Destruction, Currency

and Doubleday, 2001
Dick Foster is a longstanding McKinsey consultant and Sarah Kaplan

worked at McKinsey for many years. Foster’s thinking is influenced by his
past work on technology life cycles and by a remarkable consulting proj-
ect he did with Johnson & Johnson. The board asked Foster to help it
avoid the problems that IBM had run into in the early 1990s. Creative
Destruction suggests the following:
1 The problem that McDonald’s faces stems from failing to change as

fast as the market: failing to create new businesses and destroy old busi-
nesses fast enough to keep the momentum of value creation going.
This failure is common among large companies, especially those with
279
long periods of success where management’s attention is focused on

“operational work” rather than “adaptive work.”
2 The solution is to focus the top management team on adaptive work in
a way that does not take their eyes off the operational demands of their
current businesses. This requires them to design a process for doing
adaptive work.
3 The process for adaptive work will involve the executive group in dif-
ferent kinds of meetings with different agendas and different conversa-
tions. The people attending the meetings are likely to include more
than just the top team. Preparation, such as visits to customers or
Silicon Valley, is likely to be required so that executives come to the
meetings with a visceral understanding of the issues discussed. The key
is that the top team decide what the critical issues are and engage
wholeheartedly in figuring out how the company can respond, remem-
bering that more creation and more destruction are both likely to be
needed.
4 In addition, McDonald’s will probably need to change its strategic
planning process to encourage more divergent thinking and extended
dialogue; its R&D activities to recognize that important developments
may rightly happen outside the company; and its corporate venturing
to create a more integrated approach to new business development. All
three activities need to be focused on exploring the periphery where
new technologies and new business models are being created. Rather
than treating new business development and R&D as things that hap-
pen in a dark corner, the efforts should be central to the strategic plan-
ning debate.
5 McDonald’s will also need to change the balance between “control”
and “permission.” The current culture is likely to be killing too many
good ideas before they get aired and discouraging managers from tak-
ing risks.
COMMENTS
We like Foster’s starting point very much. He explains that most value is
created by new companies, and most of these are destroyed by the market
before they can survive for very long. The ones that do survive under-
perform the average, because the market changes too fast for companies
to keep up. In other words, the process of new companies being created
280
THE GROWTH GAMBLE
and then destroyed produces more value than that produced by companies
who survive for long periods.
We are less certain, however, about his solution, especially since he has
no examples of companies consistently outperforming as a result of using
his solution. Like so many other authors, Foster is searching for an answer
to the riddle of long-term performance, recognizing, at the same time,
that he has no exemplars. His solution is logical but probably ineffective.
His exemplar case studies are companies like Corning, J&J, and (unfortu-
nately) Enron, but it is clear that his real admiration goes to the private
equity industry.
We suspect that his admiration is well placed. The private equity indus-
try is effective at doing something that organizations set up to run a busi-
ness (like McDonald’s) can never do. We draw the parallel with corporate
venture capital. Despite three waves of attempts, corporations have
demonstrated incompetence in this area for totally predictable reasons. To
advise companies to become more like private equity firms is to advise
them to copy a game they can never win. If significant creative destruction
is needed, is it not better that the company sells itself to a private equity
firm than attempts to learn the management skills that these firms have
been honing for years?
Maybe this is what Foster should really be recommending to
McDonald’s. Selling to a private equity firm at a crucial creative/destruc-
tion moment later in a company’s life could be equivalent to using a ven-
ture capital firm to help kick start a company early in its life. We suspect
that if McDonald’s were bought by a private equity firm, the topic of new
businesses would be taken firmly off the agenda, the partner brands would
be sold, and all management’s attention would be focused on optimizing
the hamburger business.
Another part of Foster’s proposal with which we are uncomfortable is his
recommendation that companies increase “permission” and reduce “con-
trol.” This suggestion presumes that there are lots of good ideas and good
people being suppressed by the current control systems. By giving managers
more room to take risks and experiment, companies will uncover a wealth of
new value creators. We wish Foster were right. But despite many attempts by
many companies to do just what he suggests, there is no evidence that it
works. We believe that the good ideas for new businesses and entrepreneur-
ial managers in McDonald’s (and there have been examples of both) will force
their way into the limelight without much nurturing. Those that need more
permission and less control are probably not worth supporting.
281
The part of Foster’s thesis we do like, however, is his view that the new
business agenda should be firmly in the hands of the top managers.
Strategic planning, R&D, and venturing activities should not promote
1,000 flowers but rather be targeted to explore some carefully chosen
avenues. Foster appears to suggest that managers should not abdicate their
responsibility to “select” the issues and the solutions. Whereas we believe
that managers can use an intellectually rigorous selection process (the
Traffic Lights) to guide their judgments, Foster suggests that managers
should decide through debate and intuition based on a visceral exposure to
the issues.
Gary Hamel, Leading the Revolution, Harvard Business School

Press, 2000, and discussions with the authors
Gary Hamel needs little introduction: he is the world’s leading strategy guru.
“Never has incumbency been worth less. Schumpeter’s gale of creative
destruction has become a hurricane. Blink and you miss a billion-dollar
bonanza.” The solution, he believes, is to out-innovate the innovators: to
become an industry revolutionary. Radical new business concepts (busi-
ness models) come from a combination of luck and foresight, and are
driven by activists who have the imagination and drive to try the improb-
able. McDonald’s needs to start down a path of creating activists and sup-
porting radical innovations.
McDonald’s, according to Hamel, should not focus only on cost reduc-
tion or top-line growth or share buybacks or acquisitions. These may be
needed but will not provide a sustainable future. McDonald’s future will
depend on business model innovation, not only in its core business but in
new businesses around the core. Sooner or later every business model
reaches the point of diminishing returns and McDonald’s is probably well
past that point.
First, managers need to learn to think about business models. Hamel
has an impressive array of tools to help them do this and lock in advantage
that will reap high returns.
Second, managers need to focus on what is changing, on trends with
implications, and on what is not happening: anything that will help gen-
erate a fresh perspective outside the magnetic field of the current business
model. Managers then need to engage in a creative, divergent process that
generates a range of new business ideas and possible new directions.
282
THE GROWTH GAMBLE
The danger at this point is to let 1,000 flowers bloom in an undisci-

plined way. Hence some choices need to be made about general directions
of development rather than specific projects. One, two, or three new
“platforms” is plenty. Sometimes experimentation is needed in order to
uncover a new platform, but more normally a new platform can be chosen
and experimentation channeled within it. Developments outside the cho-
sen direction should not be discouraged, but it is those inside the strategy
that are likely to be retained. Incubators as vehicles for experimentation
and idea generation do not appear to work.
New business models rarely emerge fully formed. They need to be dis-
covered. This requires a capability to be built: innovation skills, innova-
tion metrics, IT for innovation, and management process for innovation,
such as innovation boards.
The key to success is to manage the investment/learning ratio of each
project. Managers often overinvest too early or fail to invest in the initia-
tives that will maximize learning.
COMMENTS
Hamel’s work on business models, what he calls business concepts, and on

where new wealth comes from—radically new business models—is
impressive and right. He parallels and probably led the thinking of
Slywotzky. Hamel’s advice is also compelling and seems tautological. In
support of it he has developed some superb tools for helping managers
think about innovation.
Hamel’s views on new businesses are more sober than his views on inno-
vation in general. He recognizes how difficult it is for companies to develop
into new areas and has concluded that incubators and undirected experi-
mentation rarely succeed. However, he is in something of a dilemma, since
his broader work on innovation emphasizes the importance of experimen-
tation and the need for a large “innovation portfolio.” This causes him to
be less clear about the process of top-down selection. He proposes that
managers should commit to general directions of development—plat-
forms—rather than to individual projects. He argues that individual proj-
ects evolve as managers discover what business model works. Hence it is
better to get commitment to a platform than a project. In this way man-
agers can learn through projects without excessive pressure, yet have the
benefit of commitment to a general direction of development.
283
In our view, individual managers play a huge role in success. The diffi-
culty of committing to a platform is that it is normally not possible to find
the unique managers who are needed to pursue the four or five different
initiatives required to move a platform forward. More normally a com-
pany may have especially talented managers for only one or two initiatives,
often in unconnected areas. Better, we believe, to back the talent that
exists than to define a platform and try to move it forward with less capa-
ble or less passionate managers.
Hamel also suggests that companies need to change the way they man-
age in order to make more room for new ideas and develop innovation
skills. He believes that few companies are good at innovation. “There are
no more than a handful of companies that have even begun to build inno-
vation systems that focus on creating a steady stream of new business con-
cepts or new rules within current concepts.” But he also recognizes that
innovation ghettos, like corporate incubators, do not succeed because they
isolate their ventures from the power structure and hence the commit-
ment they need to succeed.
In our view, McDonald’s would be foolish to attempt to turn itself into
an innovation machine. The company is dominated by operating priori-
ties, appropriately so. This does not mean that McDonald’s should spurn
innovation. It just means that innovation should have its place in its busi-
ness model rather than be positioned as a disruptive influence designed to
unseat and revolutionize the business model.
There is much overlap between Hamel’s thinking and ours. However,
we have more faith in the ability to select projects wisely using the Traffic
Lights. This makes us more comfortable with top-down processes, and
less convinced of the value of stimulating innovative activity.
Rosabeth Moss Kanter, When Giants Learn to Dance, Simon

and Schuster, 1989, and Evolve, Harvard Business School
Press, 2001
Rosabeth Moss Kanter is a distinguished professor from Harvard Business

School, a guru on strategy before most of us had heard of Gary Hamel. In
this summary, we draw on only two of her books. There are also many
insights in her other books, such as Change Masters, Frontiers of Management,
Innovation, World Class, The New Managerial Work, and her latest book,
Confidence!: How Winning Streaks and Losing Streaks Begin and End.
284
THE GROWTH GAMBLE
Kanter suggests the following:
1 “Newstreams” need to be nurtured. “New ideas can easily get lost in

the momentum of the mainstream.” Companies often have habits and
processes that actively (if not intentionally) discourage new ideas.
Moreover, efforts to develop newstreams easily lose support. Hence
the first thing McDonald’s needs to do is to commit to encouraging
new initiatives and searching out management processes that are inad-
vertently discouraging them.
2 McDonald’s needs to create the “channels” that will encourage new
ideas. “The channels (that support newstreams) need to be dug again
and again, the commitment explicitly renewed.” There is no best prac-
tice here. Each company must develop its own channels. They may be
internal incubators, increased encouragement of R&D, external ven-
ture funds, special funds, centers for creativity, incentives, planning
processes that focus on new ideas, and so on.
3 Once channels exist they should be filled with many ideas by “scout-
ing” for ideas that already exist, “coaching” people with embryonic
ideas, and “inspiring” people by example.
4 McDonald’s new ideas will come in three types. There will be a small
number of big bets that have been carefully selected. There will be a
larger number of ventures and experiments of promising, but not
proven, projects. The ventures and experiments will mainly be aimed
at supporting existing businesses or the big bets. Finally, there will be
a much larger number of incremental innovations and ideas at depart-
mental level throughout the organization.
5 Innovation requires improvisation. To aid this improvisation, top man-
agement at McDonald’s should define some directions and specific
ambitions, such as “McDonald’s for the health freak” or “another fast-
food concept that we can internationalize” (the big bets). This will pro-
vide enough guidance for improvisation teams to fire-aim, fire-aim
with rapidly developed versions.
Managing newstream projects requires a different approach to that of the

mainstream. Managers need to create a different management system
from the mainstream culture (e.g., in a venture unit) yet be ready to rein-
tegrate the two management processes as soon as possible.
285
COMMENTS
We like Kanter’s advice that top managers need to set some direction for
the search process. We agree with the need for some clarity about corpo-
rate development strategy. We differ only on the issue of timing. We
believe that clarity about direction should come after ideas have been
screened, not before. The process should start with a list of ideas, followed
by screening, then a decision is made about the direction of development
(the big bets) and in which (if any) projects to invest.
We agree that improvisation is a good word to describe what managers
do once a project starts. The initial business model is rarely the one man-
agers end up with. The journey involved in turning a good idea into a suc-
cessful business is a winding and unpredictable one.
We have more concern about Kanter’s enthusiasm for action. She
believes that more innovation is good, that the problem lies with the stul-
tifying impact of the processes needed to run the existing businesses, and
that the solution is to dig deep channels that encourage new business ini-
tiatives. We have already explained why we believe it is a mistake to advise
companies to attempt to create a parallel management system for new-
streams. We are, therefore, opposed to Kanter’s encouragement of incu-
bators and venturing units, and her advice to invest in scouting, coaching,
and inspiring. We feel that these efforts will cost a significant amount and
lead to many bad projects being sponsored. This will clutter up the agenda
and possibly cause management to lose sight of the one or maybe two
good projects that McDonald’s should be aggressively pursuing.
In her latest book Confidence! Kanter reinforces this message with an
example about the BBC’s efforts to generate new projects. The process
evolved into one where 1,000 flowers were allowed to bloom. Whereas we
see this as a problem, one that contributed to the 15% cutback and thou-
sands of redundancies announced in December 2004, Kanter is less criti-
cal, arguing that companies are only successful at finding new ways
forward when they combine top-down bets with bottom-up initiatives.
Our disagreements, however, are more about degree than principle.
We acknowledge that most companies have a large number of innovations
going on at any time, both bottom up and top down. Most of these inno-
vations are focused on the existing businesses. The big point at issue is
how many should be focused on new businesses. In our view a typical com-
pany will have three sources of new business ideas. Some come from
286
THE GROWTH GAMBLE
deliberate choices, hopefully using criteria such as the Traffic Lights.

Others come from natural extensions of innovations initiated in support of
existing businesses. Still others are generated out of innovation processes
designed to come up with new businesses. We believe that the last cate-
gory needs to be controlled more tightly than Kanter suggests. But our
real difference probably lies in the middle category: the natural extensions
of sensible innovation projects. We would like this middle category to be
put through the Traffic Lights, with the expectation that 90% of the proj-
ects will get rejected. Kanter, like Burgelman, sees this middle category as
the lifeblood of a company’s future.
Richard Leifer, Christopher McDermott, Gina Colarelli

O’Connor, lois Peters, Mark Rice, & Robert Veryzer, Radical
Innovation: How Smart Companies can Outsmart Upstarts,
Harvard Business School Press, 2000
Leifer and his co-authors were all from the Lally School of Management.
They followed a dozen radical innovations over nearly 10 years.
Leifer’s focus on radical innovation is slightly different from our focus on
new businesses. “A radical innovation has the potential to produce an entirely
new set of performance features: improvements in known features of five
times or more; or a greater than 30% reduction in cost.” Hence “radical
innovation” includes some overlap with “new businesses,” but it also includes
many projects that would be considered to be part of existing businesses.
The conclusions are the following:
1 All companies should be seeking to upgrade their ability to develop

radical innovations.
2 Radical innovations face four kinds of risks: technical, market,
resource, organizational.
3 Upgrading McDonald’s ability to “do” radical innovation involves
making changes that reduce the organizational and resource risks, both
of which are largely under the firm’s control (the technical and market
risks are driven by the project):
❏ Set up an “innovation hub” to help evaluate and incubate projects.
❏ Educate senior R&D personnel about radical innovation.
❏ Provide additional sources of funding, such as a venture capital unit
and senior managers with innovation budgets.
287
❏ Develop the capability of existing businesses and divisions to take

charge of projects as they develop and use transition teams to help
facilitate the transition.
❏ Ensure that top management has committed to “stay the course.”
4 There are seven keys to effective management of a radical innovation
project:
❏ Set the expectations of the team.
❏ Identify and track uncertainties.
❏ Develop and implement a learning plan.
❏ Adopt a resource acquisition strategy.
❏ Manage interfaces.
❏ Build project legitimacy.
❏ Get the right person for the job.
COMMENTS
For radical innovations that are part of existing businesses, we agree with
most of the points made in Radical Innovation. However, for radical innova-
tions that need to be set up as “new businesses” we have some different views.
❏ We do not think it is realistic to expect to significantly reduce the

“organizational and resource risks.” These exist or not depending on
the starting position of the host company. In our view, it is better to
screen out projects that have high organizational and resource risks
than to attempt to reduce the risks by making changes to the host com-
pany. This will lead to fewer projects passing the screen, but those that
do will have the lower organizational and resource risks that Leifer has
identified as necessary.
❏ We do not think that innovation hubs or venture capital units are
appropriate for radical innovation projects that have “new leg” poten-
tial. These incubation methods can hide the project from top managers
in a way that reduces the need for them to decide whether they are
committed or not. As a result, when a funding squeeze occurs or the
project hits a setback, it is easily dropped. Incubators and venture cap-
ital vehicles often result in a reduction in management’s commitment
to “stay the course.”
❏ We do not believe that top managers should commit to stay the course
for a general ambition like radical innovation. As we have explained
288
THE GROWTH GAMBLE
before, we believe that commitment should be withheld until a project

passes the Traffic Lights.
We should not, however, leave the impression that we disagree with much
of Leifer’s work. Moreover, one aspect of the work had a profound impact
on our thinking. Leifer and his colleagues identified and tracked the
uncertainties involved in radical innovations. This led them to the catego-
rization of technical, market, resource, and organizational risks. This
analysis is a dramatic improvement on the normal matrix of technical and
market risks because it acknowledges the effects of the host company as a
risk. It was this categorization that led us to develop our Confidence
Check, described in Chapter 9. By monitoring the degree of uncertainty
and hence degree of confidence in a number of dimensions, managers can
more objectively decide which stage gates to set and whether to continue
with the project.
Adrian Slywotzky, The Profit Zone and Value Migration,

Harvard Business School Press, 2000 & 2002, and How to
Grow When Markets Don’t, Warner Business Books, 2003
Adrian Slywotzky has written some of the most influential books on strat-
egy in the last 10 years. He is a consultant at Mercer. His books suggest
the following:
1 New wealth comes from new business models, such as Starbucks in the
coffee market. If McDonald’s wants to create a significant amount of
value outside its existing business, management will need to build on or
innovate into a new business model in some related area. “In fact, even
to revive the core, McDonald’s probably needs some new business
models.”
2 The first step is to draw a map of the value patterns in the hamburger
industry and related industries. It is important to do this analysis objec-
tively and thoroughly, because organizational memory may screen
McDonald’s managers from seeing the new value areas.
3 The next step is to choose some markets/products that are in related
areas and where the business model is in the expanding and growing
stage, what Slywotzky calls the “inflow phase.” This can be done by
289
innovating a new business model or acquiring one that is in early devel-

opment. One of the most important areas in which McDonald’s can
look for these new business models is “addressing the hassles and issues
that surround the product”: additional services for customers or sup-
pliers or channels of distribution. Slywotzky refers to this form of
development as “demand innovation.” He argues that it generates bet-
ter opportunities than product innovation.
4 The third step is to build a robust business design starting with the cus-
tomer, building on business design patterns that are common
(Slywotzky lists the common ones), exploiting the company’s assets,
taking care to acquire the needed new competencies, and thinking
about the “hidden liabilities” that the company already has that may get
in the way.
5 Finally, new business designs should be kept separate from the core
until they have proved themselves and can withstand challenges from
existing businesses. At the same time they should get support from sen-
ior managers to ensure they are resourced appropriately and fully
develop their opportunity.
COMMENTS
We like Slywotzky’s top-down approach. Like him, we believe that better

top-down strategic thinking can go a long way. We support his advice about
defining the value maps and exploring the business models adjacent to the
current businesses. We like his use of the term “hidden liabilities” and the
importance he gives to this bit of organizational reality. We also like the fact
that Slywotzky does not seem to believe that success is a numbers game.
However, we are not confident that companies can easily execute busi-
ness models with which they are not familiar. In fact, we believe that “hid-
den liabilities” are often hard to get around. Rather than presuming that
there is a solution to hidden liabilities, we presume that they should be
part of the screening criteria. Companies should only approve projects
where the “hidden liabilities” are small or manageable. In the coffee mar-
ket, P&G’s coffee division would have had many hidden liabilities if it had
entered the coffee shop business. Even if P&G had identified the oppor-
tunity in time to do something about it, we would not have expected P&G
to have been an effective competitor to Starbucks.
290
THE GROWTH GAMBLE
Slywotzky’s main selection criterion is size. He encourages companies

to aim for the largest opportunities with the best ideas. “If the idea is good
enough, managers can be bolder about taking on something less familiar.”
In contrast, our Traffic Lights give little attention to the size of the profit
pool. We argue that criteria such as “size of the contribution” and “size of
the learning costs” (our term for hidden liabilities) are more important.
We suggest that companies may need to reject all opportunities and
decide to stick to the existing business. Slywotzky would consider this to
be “sticking one’s head in the sand.” Moreover, unlike Zook, Slywotzky
thinks that there is an equal danger of overinvesting in the core business.
He appears to believe that companies have almost a moral duty to try new
business models. “If not, it is bad for them, bad for their shareholders and
bad for the economy.” We are particularly concerned about this moraliza-
tion of the growth gamble. Managers who believe that they must grow are
more likely to harm themselves, their shareholders, and the economy than
managers who understand the gamble they are taking on and when big
bets are likely to succeed.
Michael Tushman and Charles O’Reilly, Winning through

Innovation, Harvard Business School Press, 1997
Tushman and O’Reilly are professors at Harvard and UCLA, spanning the
East and West coasts of America and the disciplines of organization and
business policy. They have done extensive research on the evolution of
industries and how companies cope with technological change.
“To succeed both today and tomorrow, managers must play two differ-
ent games simultaneously. First, they must continually get better at com-
peting in the short term … Managers must also master another game:
understanding how and when to initiate revolutionary innovation and, in
turn, revolutionary organizational change.”
It is hard, they suggest, to drive performance in the existing business while
developing the skills to succeed in the next-generation technology or business
model. Most fail, like RCA in the move from vacuum to solid state, Swiss
watch manufacturers in the move from mechanical to quartz, and Danish hear-
ing aid company Oticon in the move from behind-the-ear to in-the-ear tech-
nology. McDonald’s will need to be particularly careful as the industry moves
from the traditional hamburger business model to the fast-casual model.
291
During times of change, there is a period of rapid innovation when dif-

ferent technologies and business models compete for dominance. Then,
once a dominant new model emerges, there is rapid growth and huge suc-
cess for the company that can develop the skills, organization, and culture
that suit the new model. Companies wedded to an existing model nor-
mally find it particularly hard to commit fully to a new model, especially
when the new model has rules of thumb or values that conflict with the
previous way of doing things.
The first requirement is that McDonald’s runs its existing business
well. “If an organization is not successful today, there can be no tomor-
row.” This involves:
❏ Establishing a clear strategy, objectives, and vision.

❏ Identifying gaps between current performance and goals.
❏ Developing an effective problem-solving process for dealing with the
gaps.
❏ Building a culture that provides strong social control and includes
norms that support innovation—“creativity” and “implementation.”
The second requirement is to develop the ability to manage three innova-

tion streams simultaneously—incremental, architectural, and discontinuous.
❏ Incremental can be led by the existing businesses.

❏ Architectural and discontinuous innovations need separate units (one
for each project) with the freedom to break the corporate rules.
❏ Senior managers need to be comfortable with multiple competing
innovations during periods of “ferment” before dominant new business
models emerge.
❏ Vision, team diversity, and good process can be used to keep innova-
tion units separated organizationally as well as integrated into the cor-
porate strategy
❏ Top managers must be ready to “shape the closing of” a dominant new
model (i.e., decide when the organization should switch from experi-
menting with to driving a new model).
❏ The top management team must prepare themselves so that they are
ready and able to lead change.
292
THE GROWTH GAMBLE
COMMENTS
We agree with almost all of Tushman and O’Reilly’s observations about

organizations, organizational inertia, and the difficulty of achieving align-
ment around a new business model or innovation when the organization
is already committed to an existing one. We also agree with them that
many innovations, and hence all new businesses, require a different com-
bination of mindset, rules of thumb, norms, people, and structures. Their
book superbly explains the challenges that new businesses pose. It is
matched only by Christensen in its understanding of the realities of
organizations.
We also applaud Tushman’s advice that companies should focus on put-
ting their existing house in order before they try looking for new things.
We part company, however, on the practical issue of what is possible
and on the wisdom of developing a vision before the opportunity for a new
“dominant model” becomes clear.
Tushman argues that vision provides one of the critical integrating
mechanisms that make it possible for companies to experiment in sepa-
rated units while keeping some integrated approach to innovation. They
are right, but only if top managers are capable of defining a wise vision,
before the innovation starts.
Our observation is that this is rare. There are many visions: Intel’s
“internet” vision, McDonald’s “partner brands” vision, AT&T’s “comput-
ers and communications” vision. The problem is that these visions are
often more of a hindrance than a help. Our view is closer to that of Lou
Gerstner, who decided not to have a vision for IBM, at least until the path
forward was clear. This may be one of the reasons Tushman comments:
“Building ambidextrous organizations and initiating revolutionary change
is difficult … only a small minority of farsighted firms initiate discontinu-
ous change before a performance decline.”
Our recommendation is different. We suggest that managers review
the new business opportunities, choose one or two that pass the Traffic
Lights, and invest in them based on the strength of the Traffic Lights logic
rather than some predefined vision. As success develops, so also should a
new vision. In this respect our view is closer to that of Burgelman. The
new corporate strategy should, we believe, emerge as projects succeed.
On the issue of practicality, we have not observed many managers who
can handle the ambidextrous organizational requirements that Tushman
293
proposes. For many, creating a cohesive top team is a huge challenge, at

least until the new direction has become clear. For others, the ability to
maintain competing experiments in separated units and yet integrate them
into the corporate-level strategy is too difficult. In fact, Tushman’s
research demonstrates that the ambidextrous ambition is too difficult for
most management teams.
The alternative in our view is to acknowledge the weaknesses: mindset
limitations, cultural constraints, and other restrictions of the existing
organization and top management. These limitations and constraints can
then be used to help screen opportunities. They can be used as part of the
Traffic Lights under the headings of “our contribution,” “learning costs,”
and “leadership/sponsorship.” New businesses that are difficult because
they clash with the constraints are eliminated. Sometimes, especially when
the constraints are substantial, all businesses will be eliminated. In this
case, focusing on the core or designing projects that eliminate some of the
constraints, for example through an acquisition, can be more useful than
trying to develop the capability to be an ambidextrous organization.
Like Christensen, Tushman and O’Reilly have produced a superb diag-
nosis of the problem. But their solution is more theory than practice.
There are so few companies demonstrating the capabilities they claim are
required that we believe a less ambitious solution is more practical.
Chris Zook & James Allen, Profit from the Core, Harvard
Business School Press, 2001 and Chris Zook, Beyond the Core,
Harvard Business School Press, 2004
Chris Zook and James Allen are consultants at Bain & Co. Their work
suggests the following:
1 McDonald’s needs to focus on its core business and make sure that it is
continuing to dominate its industry and produce good financial results.
This work may involve some debate about what the core business is.
For McDonald’s the debate is unlikely to be very protracted, and given
the current performance problems it is best to define the core narrowly,
at least initially.
2 Careful analysis is needed of the causes of the current hiccup in the
core business and appropriate action to be taken. This may require
294
THE GROWTH GAMBLE
investments in new products, improved premises, added services, and

so on. Revitalizing the core may require significant innovative behav-
ior. The greatest performance improvements, however, are likely to
come from growing McDonald’s strongest geographic businesses
rather than from turning around its weaker businesses. Paradoxically, it
is the strongest businesses that are usually underperforming the most.
3 While work is continuing on the core, management should do a full
potential analysis of the core business, looking at all the growth oppor-
tunities that this strong core provides. It is likely that these opportuni-
ties will offer McDonald’s good potential. Managers often
underestimate the potential for growth from the core.
4 When the core is running smoothly again and plans to exploit its full
potential are being executed, McDonald’s may want to consider adja-
cent opportunities. The Bain list of adjacency dimensions suggests that
the best prospects might lie in backward integration or in using
McDonald’s capabilities in similar food formats:
❏ new geographies: part of the core;
❏ new value chain steps: not obvious but if found would be part of the
core;
❏ new channels: not obvious, e.g., hamburgers in supermarkets
unlikely to be very profitable;
❏ new customer segments: part of the core unless it involves a new
retail format;
❏ new products: part of the core unless it involves a new retail format;
❏ backward integration: may offer opportunities;
❏ forward integration: not obvious;
❏ capabilities: most likely if using current fast-food skills in new for-
mats.
5 The list of adjacencies can then be screened against criteria such as
impact on core, potential to become a leader, robustness of the profit
pool, defensive value, potential for further developments, probability
of executing well, and strength of relatedness to the core (number of
adjacency steps from the core). McDonald’s should be cautious.
Adjacency expansion is risky. “The most successful companies at adja-
cency expansion have some of the most restrictive criteria.”
295
COMMENTS
Zook’s thesis, like Peters’ “stick to your knitting” before it, is well
grounded in fact. We like the emphasis on getting existing businesses into
winning market positions before spending time on developing new busi-
nesses. We also like Zook’s focus on doing a “full potential” analysis of the
core business. This helps ensure that companies do not lose faith in their
core too early. The fact that Zook’s sustained value creators nearly all had
strong core businesses is a powerful incentive for companies to try to suc-
ceed in one or two businesses before pushing their luck in new areas.
However, Zook does not fully address the very real problem facing many,
in fact most, companies: they are number two or three or worse in their exist-
ing businesses with little prospect of becoming number one. Hence they are
motivated to find something else to work at or some adjacent business to link
with that will help them compete. Our solution to this problem is not to
actively discourage managers from considering new businesses just because
their core is weak, but to equip them with a powerful screening tool—the
Traffic Lights—to help them avoid doing anything stupid.
We like Zook’s concern that new business efforts may distract man-
agers from the core. We find that managers frequently underestimate the
distraction risks.
Zook’s focus on adjacencies is also laudable. His definition of what is or
is not an adjacency, however, is not very enlightening. An adjacency is dif-
ferent from a diversification to “the extent to which it draws on the cus-
tomer relationships, technologies or skills in the core business to build
advantage.” In other words, the company must have something to bring to
the party that is important. This is paragraph 1.1 of most books on new
businesses, hence it does not add much to common sense. Nevertheless,
Zook’s adjacency maps do help managers identify a large number of
potential adjacencies and his concept of number of adjacency steps away
from the core is a practical tool for assessing potential learning costs.
Recognizing that many companies are “awash with adjacencies,” Zook
provides a list of screening questions. Most of these questions overlap to
some degree with parts of the Traffic Lights. In fact, Zook reviewed our
Traffic Lights and commented that they are almost identical to the crite-
ria he uses.
However, on close reading of his books we believe that the Traffic
Lights are an advance on Zook’s thinking in a number of ways. He gives
296
THE GROWTH GAMBLE
examples of success—Dell into servers—and of failure—Quaker Oats into

Snapple—but does not show how their screening process would help man-
agers decide whether a new opportunity is a “server” or a “Snapple.”
Moreover, Zook does not suggest, as we do, that the screening process
may screen out all of the identified adjacencies. His message implies that,
once McDonald’s has sorted out its core business, there are likely to be
plenty of adjacencies to pursue. Our experience suggests that there are
plenty of adjacencies, but few that will pass the Traffic Lights.
Overall, we feel that Zook’s ideas are the closest to our own and fit our
data better than the ideas of the other major authors.
297
APPENDIX B
A DATABASE OF
SUCCESS STORIES
This appendix describes the database of success stories we have identified

as part of the research. It is not a statistically representative database.
Rather, it is a collection of interesting examples of success that we found
in the course of our work.
We wanted to create a database of examples of companies that had suc-
cessfully created a significant new business. The definitions of “business,”
“new,” “significant,” and “successful” were:
❏ A separate organization unit rather than merely a new product, chan-

nel, or market.
❏ A business model new to the parent and/or requiring new parenting
behaviors.
❏ Sales or profits amounting to 20% of the parent company total or
shareholder value of $1 billion (whichever was the smaller number).
❏ Considered at the outset to be a permanent part of the portfolio, and
successful enough to generate a return on investment equal to or
greater than the cost of capital.
As shown in Figure B.1, we have built up information on 66 cases. Some

have ended up falling well short of the criteria, even proving considerable
failures, like the move of Lattice (when BG Group) into telecoms. Others
are better classified as potential or modest successes. Of the 66 cases, 54
meet our criteria to a greater or lesser degree. Of these we have inside
information through company interviews on 58% and publicly available
information on the remaining 42%.
Of the 54 successes, 9 (17%) have been sold or floated wholly or par-
tially. However, they are included because at the outset the intent was for
THE GROWTH GAMBLE
Figure B.1 Success levels and information quality in the database

these businesses to become significant additions to the portfolios of their
parent companies.
THE SUCCESSES
Some companies, like Dixons and Whitbread, have had a number of suc-
cesses. So we have 54 examples from 44 different companies, as listed in
Box B.1 overleaf.
In addition to these examples we researched seven companies that have
repeatedly created new businesses. These serial developers are 3M,
Canon, CP Group, CRH, ServiceMaster, Thermo Electron, and Virgin.
Because each of these companies has created many new businesses, we did
not include any individual examples in our database.
LOCATION AND MATURITY OF DEVELOPMENTS
Our sample is weighted toward US and UK companies (Figure B.2). It

spans over 20 years (Figure B.3), with examples covering a spread of
maturities.
299
A DATABASE OF SUCCESS STORIES
Figure B.2 Location of parent companies
Figure B.3 Timing of initial major commitments
300
THE GROWTH GAMBLE
Box B.1 The database of success stories
Corporate parent Nationality Development

ABF UK British Sugar; a move into sugar refining and
distribution through acquisition.
Acer Group Taiwan PCs; the creation of a major PC business from a
core that produced microprocessors for games
manufacturers.
Anglian Water UK Anglian Water International; the development
Group of an international business following
deregulation in the UK.
Anglo American South Africa Mondi Europe; move into European forest
products, from a core business in minerals and
forest products in South Africa
Barclays Bank UK Barclaycard; move into credit card business (first
in the UK) in the late 1960s from commercial
banking.
Bombardier Canada Rail cars for mass transit systems; expansion
from snow-going equipment for industrial and
commercial use (snowmobiles) in 1974.
Bombardier Canada Aerospace (regional aircraft and executive jets);
move in 1986 through the initial acquisition of
Canadair.
Boots UK Boots Healthcare International; development of
international over-the-counter (OTC) self-
medication business (Nurofen, Optrex etc) from
remnants of sale of pharma business. Core
business is drug stores.
Cardinal Health US Pharmaceutical products distribution; a major
shift from initial food distribution through
acquisition and divestment, and then growth
through geographic expansion, followed by
forward and backward integration, and moves
into adjacent products.
Centrica UK Electricity supply; move by British Gas, the UK's
leading gas retailer, to sell electricity post-
deregulation.
301

Centrica UK Roadside services;diversification through
acquisition of the AA,the UK's leading
automobile emergency and support organization.
Charles Schwab US OneSource; development of the mutual fund
supermarket, and subsequent transformation to
become a leading e-broking firm.
Diebold US ATMs; move into ATMs and ancillary services in
the early 1970s by a company then specializing
in storage, security and record handling equipment.
Dixons UK Freeserve; development of an internet service
provider by UK electrical and photographic
equipment retailer.
Dixons UK PC World; development of UK's largest chain of
computer superstores from its inception in 1991.
Dixons UK The Link ; development of a major UK chain to retail
mobile phones.
DSM Netherlands Chemicals; switch by major Dutch mining company
out of mining into chemicals.
DSM Netherlands Life sciences; move by DSM into to life sciences
(pharmaceutical ingredients) leading to decisions to
exit parts of the chemicals business.
EMAP UK Radio stations; move by UK magazine publisher into
radio.
GE US GE Capital Corporation; creation of one of the largest
global financial services companies by diversified
manufacturing company
GE US NBC; major move into television through acquisition
of NBC as part of larger acquisition of RCA in 1986.
Gillette US Braun; move into small domestic electrical
appliances through acquisition in 1967.
GrandMet UK Move by hotel and restaurant business into drinks
through acquisition in mid-1970s of IDV as part of
Trumans deal.
GPS France GrandVision; transformation of GPS, a French
business in one-hour photo processing, into
GrandVision, a company focusing on instant and
discount optical services.
302
THE GROWTH GAMBLE

HDFC India HDFC Bank; development of an Indian retail and
commercial bank by India's leading specialist
housing company.
Hewlett-Packard US Computers; HP's transition into a major
computing company (emphasizing post-1980
developments).
Hutchison Whampoa China Orange; creation of a major UK mobile phone
company prior to its sale to Mannesman in 1998.
IBM US IBM Global Services; creating the world's largest
computer services and consulting business,and now
IBM' s biggest business.
IBM US IBM PC; creation of IBM’s PC business.
Irish Life Ireland Irish Life & Permanent; move by Irish Life, Ireland's
leading life insurer, into retail banking through
merger with Irish Permanent Building Society in
1999, and subsequent acquisition of Irish TSB
Bank in 2000.
Johnson Matthey UK Johnson Matthey Pharmaceutical Materials Division;
the development of a significant fourth leg of the
specialty chemicals company.
Kelloggs US Nutrigrain; entry into the convenience food business
in 1970s.
Kelloggs US Keebler; creation of a major convenience food
business through acquisition of Keebler, a better
vehicle for its own convenience products.
Marks & Spencer UK M&S Financial Services; move by major UK retailer
into financial services.
Maryland National US MBNA; creation in 1982 of an in-house credit card
Bank business, prior to its float in 1990 to become one of
the largest international credit card companies on
the basis of serving affinity groups.
Midland Bank UK First Direct; a separately branded direct telephone
bank launched in 1989 (now part of HSBC).
Neptune Orient Singapore APL Logistics; transformation of a division of
Lines (NOL) American President Lines (APL), acquired by NOL in
1997, into a global origin-to-destination logistics
company.
303

Nokia Finland Mobile phones; transformation of a broadly
diversified company spanning rubber products to
electronics into a specialist mobile phone company.
Philips Netherlands Polygram; creation of a global recorded music
company by Europe’s leading consumer electronics
company
Philips Netherlands Navigation Technologies (Navteq); creation of a
digital map database company.
Prudential UK Egg; creation in 1998 of the UK internet bank and
credit card company by Prudential, one of the UK's
largest insurers.
Rentokil UK Office plants and other services; move by company
specializing in vermin control into wide range of
businesses providing office support services.
Reuters UK Reuters Monitor Dealing Service;transformation of a
rather non-commercial but prestigious news agency
in 1963 to a major provider of global electronic
information and trading services to the financial
sector by 1981.
Rexam UK Metal packaging business; entered metal packaging
as part of a transformation of the widely diversified
former Bowater Group through focus on consumer
goods packaging.
Royal Bank of UK Direct Line; creation of the UK's biggest direct-
Scotland selling, general insurance company through an
initial move into motor insurance.
Seagram Canada Media and entertainment;entry of drinks group into
media and entertainment, primarily through
acquisition of MCA, owner of Universal Studios, in
1994 and Polygram in 1995,prior to subsequent sale
to Vivendi in 2000.
South Staffs Water UK HomeServe; creation in 1991 of a domestic
emergencies assistance and insurance business by a
UK water utility.
SunAmerica US Retirement savings; transformation of a mortality-
based life insurance company into one specializing
only in the retirement savings market.
304
THE GROWTH GAMBLE

TRW US TRW Automotive;major new business concentrating
on automotive safety developed in early 1990s to
replace declining space and defense activity (since
partially floated by Northrop Grumman Corp).
VNU Netherlands Marketing services; transformation of a Dutch
publishing company into a major international
supplier of business information.
Whitbread UK Restaurants; entry of UK pub chain into restaurant
business, including Beefeater chain.
Whitbread UK Travel Inn; use of Beefeater restaurant locations to
set up a budget hotel chain.
Whitbread UK Hotel business; initial acquisition of Swallow Hotels
transformed through a further acquisition and
acquiring the UK Marriott franchise, such that the
hotel business has become the largest business in
the group.
WPP UK Media and communication services; an expansion
from the original core of advertising services.
305
NOTES
CHAPTER 1: THE CHALLENGE OF NEW BUSINESSES
1 Robert Burgelman, Strategy Is Destiny, Free Press, 2002.

2 Mats Lederhausen believes that this report was inaccurate: that
McDonald’s never intended to sell a stake in these businesses.
However, the report illustrates the shift in attention that was taking
place at McDonald’s when Jim Cantalupo took over.
3 Gary Hamel refers to this problem in a slightly different way. He
observes that failure with new initiatives is often due to the invisible,
untested assumptions that executives make about the environment and
appropriate business formula. These assumptions are easy to get wrong
because they are the ones made about and appropriate for the existing
businesses; whereas our solution is to bias the search for new businesses
toward situations where these invisible assumptions will not be a dis-
advantage. Hamel believes that, by being exceptionally clear about
what is known and what is not known, it is possible to manage a learn-
ing process that avoids the pitfalls of these untested assumptions.
4 Business Week, May 3, 2004.
5 Jay Greene, Microsoft’s Midlife Crisis, Business Week, April 19, 2004.
6 There is a third way for companies to get into new businesses that does
not involve a crisis or finding businesses that fit. Philips’s success with
Polygram is an example. If the business is far enough removed from the
core that there is no attempt to apply the thinking of the existing busi-
nesses to the new area, if the management team are given sufficient
autonomy, and if there is still a strategic logic that makes sense, success
is possible. However, it normally results in the business being sold
because it cannot be integrated into the parent company portfolio.
Hence it does not really meet our criteria of developing new growth
avenues for a company with a declining core. It is also comparatively
rare.
THE GROWTH GAMBLE
CHAPTER 2: BEATING THE ODDS
1 Corporate Strategy Board, Growth Restarts: Reinvigorating Principled

Revenue Growth in Mature Companies.
2 Chris Zook and James Allen, Profit from the Core, Harvard Business
School Press, 1999.
3 Gary Hamel (Leading the Revolution) gives particular attention to the
processes of idea generation, incubation, and “early testing.” Mehrdad
Baghai (The Alchemy of Growth) suggests improvements to incubation
processes as a way of filling the pipeline. Rosabeth Moss Kanter (When
Giants Learn to Dance) recommends improvement to incubation,
launch, and business development processes. Clayton Christensen (The
Innovator’s Dilemma) encourages managers to focus on disruptive tech-
nologies and greater separation of new businesses as they are being
developed. Robert Burgelman (Strategy Is Destiny) calls for improve-
ments in bottom-up processes, the role of middle managers, and the
ability of corporate-level managers to suspend judgment. Richard
Foster (Creative Destruction) focuses on the need for richer dialogue
between the senior executives at the top of companies.
4 Andrew Campbell, Julian Birkinshaw, Andy Morrison and Robert van
Basten Battenberg, The Future of Corporate Venturing, Sloan
Management Review, Fall 2003.
5 Chis Zook and James Allen, Profit from the Core, Harvard Business
School Press, 1999.
6 Jim Collins and Gerry Porras, From Good to Great, HarperCollins,
2001.
7 In alphabetical order, Mehrdad Baghai, The Alchemy of Growth; Zenas
Block, Corporate Venturing; Robert Burgelman, Strategy Is Destiny;
Clayton Christensen, The Innovator’s Solution; Gary Hamel, Leading
the Revolution, Richard Foster, Creative Destruction; Richard Leifer,
Radical Innovation; Rosabeth Moss Kanter, When Giants Learn to Dance
and Evolve!; Adrian Slywotzky, How to Grow When Markets Don’t;
Michael Tushman, Winning through Innovation; Chris Zook, Beyond the
Core.
307
NOTES
CHAPTER 3: THE DIFFICULTIES OF BUILDING NEW LEGS
1 Clayton Christensen and Michael Raynor, The Innovator’s Solution,

Harvard Business School Press, 2003.
2 Richard Foster and Sarah Kaplan, Creative Destruction, Currency, 2001.
3 Value destruction here is defined by the cost of capital. Any company
that performs less well than the average is destroying value, because the
cost of equity is calculated based on the average return that share-
holders expect from investments in the capital markets.
4 BG demerged into Lattice and BG Group. Lattice subsequently
merged with National Grid and the remaining part of the telecoms
activity (SST) was integrated into National Grid’s telecoms business
(Gridcom). This formed a platform from which National Grid
acquired Crown Castle, to become a market leader in telecoms towers
in the UK. At the time of writing it is not clear whether this new busi-
ness for National Grid Transco (the new merged company) is a success.
Moreover, it is clear that the attempts by Lattice to build a towers busi-
ness (SST), although not a financial disaster, were not viewed by man-
agement as a success. Within BG Group, two or three of the original
new businesses continue with limited support and the main opportu-
nity, domestic combined heat and power, has still not developed a tech-
nology at a low enough price point to be commercial.
5 Michael Porter, Competitive Strategy, Free Press, 1983.
6 Igor Ansoff, Corporate Strategy, McGraw-Hill, 1965.
7 Gary Hamel and CK Prahalad, The core competence of the corpora-
tion, Harvard Business Review, May 1990.
School Press, 1999.
9 Careful reading of Hamel, Kanter, Slywotzky, and other authors does
not reveal sections where they expressly recommend companies to
“play the numbers game.” Yet the idea that companies need to multi-
ply the number of experiments and ventures pervades their texts.
When pushed, most authors reject the “1,000 flowers bloom” strategy,
yet encourage companies to be less deterministic, less risk averse, and
more experimental.
10 Gary Hamel, Bringing Silicon Valley inside, Harvard Business Review,
May 2000.
11 Julian Birkinshaw, Entrepreneurship in the Global Firm, Sage, 2000.
308
THE GROWTH GAMBLE
12 In our discussions with consultants at Strategos and Bain & Co., we

were told of research projects both firms had done that confirmed the
lack of success of corporate venturing units. It should be noted that
there is other research that is more favorable to corporate venturing
units. A good summary of this research is contained in Corporate
Venturing by Zenas Block and Ian MacMillan, Harvard Business School
Press, 1993. Our explanation for the contrasting views on corporate
venturing is given in Chapter 8.
CHAPTER 4: WHEN LOW GROWTH IS BETTER THAN

GAMBLING
1 With growth at around 2–3% and a cost of capital of 12–15%, market

capitalization increases by 10% for each 1% increase in growth, using
the dividend growth model for valuation. If returns are far in excess of
the cost of capital, each 1% increase in growth can have a much larger
(25%) impact on market capitalization. As the growth rate approaches
the cost of capital, the dividend growth model produces very large
valuations.
2 Corporate Strategy Board, Stall Points, Corporate Advisory Board,
1998.
3 Arie de Geus, The Living Company, Nicholas Brealey Publishing, 1999.
1998.
5 Lou Takes the Gloves Off, Business Week, November 18, 2002.
6 David Garvin, Emerging Business Opportunities at IBM (A), Harvard
Business School, Case No N9-304-075.
1998.
CHAPTER 5: THE NEW BUSINESSES TRAFFIC LIGHTS
1 After the manuscript was complete, Chris Zook commented that

“Figure 5.3 is almost exactly what evolved from our analysis of 12
paired company comparisons where we examined over 500 growth
moves one by one and identified the key thematic differences that
could have been assessed better by research.” As we point out later, the
309
NOTES
Traffic Lights are not built on dramatic new theory. There are some
new insights and some old ideas presented in new ways. But the main
contribution of the Traffic Lights is to pull together the excellent work
that has been done by strategy academics over the last 30 years and
practitioners such as Zook and Allen at Bain & Co.
2 Michael Porter, Competitive Advantage, Free Press, 1986.
3 Richard Foster and Sarah Kaplan, Creative Destruction, Currency
Doubleday, 2001.
4 Barry Gibbons, If You Want to Make God Really Laugh Show Him Your
Business Plan, Capstone, 1998.
5 Gary Hamel, Waking up IBM, Harvard Business Review, July–Aug 2000.
6 James Mackintosh, Poor quality of Mercedes cars dents profits,
Financial Times, Oct 29, 2004.
CHAPTER 6: DIVERSIFICATION
1 Tom Peters and Robert Waterman, In Search of Excellence, Free Press,

1984.
2 Igor Ansoff, Cororate Strategy, New York: McGraw-Hill, 1965.
3 Richard P. Rumelt, Strategy, Structure and Economic Performance,
4 Gary Hamel and CK Prahalad, The Core Competence of the
Corporation, Harvard Business Review, May 1990.
5 Michael Goold, Andrew Campbell, and Marcus Alexander, Corporate-
Level Strategy, John Wiley & Sons, 1994.
6 Ibid., pp. 340–53.
CHAPTER 7: SEARCHING FOR NEW BUSINESSES
1 Mehrdad Baghai, Stephen Coley, and David White, The Alchemy of

Growth, Free Press, 1998.
2 Gary Hamel, Leading the Revolution, Harvard Business School Press, 2001.
3 Rosabeth Moss Kanter, When Giants Learn to Dance, Harvard Business
School Press, 1988 and Evolve!, Harvard Business School Press, 2000.
5 Clayton Christensen, The Innovators’ Solution, Harvard Business School
Press, 2003.
310
THE GROWTH GAMBLE
School Press, 1999.
8 Julian Birkinshaw, Entrepreneurship in the Global Firm, Sage, 2000.
9 Rosabeth Moss Kanter (When Giants Learn to Dance is the author
who gives most emphasis to the importance of having a parallel process
for developing new businesses. She distinguishes between processes for
“mainstream” businesses and processes for “newstream” businesses.
She is in good company: most other authors agree with her.
10 In a letter to Harvard Business Review in response to Gary Hamel and
Gary Getz’s article Innovating in an age of austerity, Harvard Business
Review, July 2004.
11 Michael Goold, Andrew Campbell, and Marcus Alexander, Corporate-
Level Strategy, John Wiley & Sons, 1994
12 Robert Burgelman, Strategy Is Destiny, Free Press, 2002, pp. 368–79.
13 Ibid., p. 368.
14 Clayton Christensen also recommends that managers need to be ready
to cut failing projects. His advice, with regard to disruptive
innovations, is to be “impatient for profit and patient for sales.”
Isabelle Royer (in Why bad projects are so hard to kill, Harvard
Business Review, February 2003) explains why it is so hard to cut proj-
ects and offers some remedies: beware of cheerleading squads, establish
an early warning system, and recognize the role of the exit champion.
15 David Garvin and Lynne Levesque, Emerging Business Opportunities
at IBM (A), Harvard Business School Case No N9-304-075.
16 Robert Burgelman, Strategy Is Destiny, Free Press, 2002, p. 373;
Rosabeth Kanter, When Giants Learn to Dance, Harvard Business
School Press, 1989.
CHAPTER 8: IS THERE A ROLE FOR CORPORATE

VENTURING?
1 These labels were first used in an article in Sloan Management Review

titled The future of corporate venturing, Andrew Campbell, Julian
Birkinshaw, Andrew Morrison, and Robert van Basten Battenberg,
Sloan Management Review, Fall 2002.
2 See H. Chesbrough, Making sense of corporate venture capital,
Harvard Business Review, March 2002, pp. 4–11. See also R. A.
311
NOTES
Burgelman, Designs for corporate entrepreneurship in established

firms, California Management Review, vol. XXVI, no. 3, 1984, pp.
154–66; and Strategy Is Destiny, Free Press, 2002. Rosabeth Moss Kanter
has also produced a number of important case studies published either
in the Journal of Business Venturing or the Harvard Business School case
series. These cases are discussed in her books from Harvard Business
School Press, When Giants Learn to Dance (1988) and Evolve! (2000).
3 Lucent New Ventures Group is written up in detail by Rosabeth Moss
Kanter and M. A. Heskett, Harvard Business School Cases 9-300-085,
9-601-102, and teaching note 5-301-108, also by Henry Chesbrough,
Designing corporate ventures in the shadow of private venture capital,
California Management Review, vol. 42, no. 3, 2000, pp. 31–49.
4 Coller Capital focused on the “secondary” market in which entire port-
folios of investments are traded, typically at a substantial discount.
Lucent retained a minority stake when it sold NVG to Coller.
5 More specifically, Intel Capital makes four different types of invest-
ments: (1) ecosystem investments, to increase demand for the micro-
processor, (2) market development investments, such as Ariba Online,
to help accelerate technology adoption in a foreign country, (3) gap
fillers, which are investments intended to fill a gap in a growing mar-
ket, and (4) eyes and ears investments, which are targeted toward new
disruptive technologies that Intel might otherwise miss.
6 The idea of separating out certain activities from their normal func-
tional management is well established in the literature. See P. F.
Drucker, Innovation and Entrepreneurship: Practice and Principles, Harper
and Row, 1985, and Clayton Christensen, The Innovator’s Dilemma,
7 Christensen (1997) argues that disruptive technologies need to be
managed through a separate business unit, though subsequent research
has suggested that the decision is contingent on a number of different
variables. Venture innovation can be a highly effective way of making
the initial investment in a disruptive technology, but we would expect
the technology in question to be transferred to a business unit once its
commercial potential is recognized.
8 Gifford Pinchot coined the term in his book Intrapreneuring, Harper &
Row, 1985.
9 The foremost proponent of this idea was Gary Hamel, but many other
consultants and academics made similar arguments. See G. Hamel,
312
THE GROWTH GAMBLE
Leading the Revolution, Harvard Business School Press, 2001, and R.

Foster and S. Kaplan, Creative Destruction, Doubleday, 2001.
10 It is worth noting that the statistical evidence collected in the London
Business School questionnaire showed that venture innovation often
fails because of excessive autonomy. That is, there was a strong, statis-
tically significant correlation between venture unit performance and
operating autonomy, and between venture unit performance and oper-
ating budget independence.
11 The success of GE Equity up to 2000 is well documented in Don
Laurie, Venture Catalyst, Nicholas Brealey Publishing, 2001.
12 See M. Maula and G. Murray, Corporate venture capital and the cre-
ation of US public companies: The impact of sources of venture capi-
tal on the performance of portfolio companies, in M. A. Hitt, R. Amit,
C. Lucier, and B. Shelton (eds) Strategy in the Entrepreneurial
Millennium, John Wiley & Sons, 2001.
13 See G. Hamel, Leading the Revolution, Harvard Business School Press,
2001, and C. Markides, All the Right Moves, Harvard Business School
Press, 2000.
14 Ralph Biggadike studied corporate diversification in the 1970s and
concluded that a major cause of failure was management’s timidity. See
R. A. Biggadike, The risky business of corporate diversification,
Harvard Business Review, May–June 1979; or R. A. Biggadike, Corporate
Diversification, Harvard Business School Press, 1979.
CHAPTER 9: POSITIONING AND SUPPORTING A NEW

BUSINESS
1 Zenas Block and Ian MacMillan, Corporate Venturing: Creating New

Businesses within the Firm, Harvard Business School Press, 1993.
2 Clayton Christensen, The Innovator’s Dilemma, Harvard Business
School Press, 1997.
3 Charles O’Reilly and Michael Tushman, The Ambidextrous
Organization, Harvard Business Review, April 2004.
4 Clayton Christensen and Michael Raynor, The Innovator’s Solution,
Harvard Business School Press, 2003, Chapter 7.
5 Robert Burgelman and Leonard Sayles, Inside Corporate Innovation,
Free Press, 1986.
313
NOTES
6 Tushman, Smith, Wood, Westerman and O’Reilly also emphasize the

importance of separation in Innovation Streams and Ambidextrous
Organization Designs, submitted to Academy of Management Journal,
July 2004. They provide compelling data on the advantages of
separation.
7 Michael Goold and Andrew Campbell, Designing Effective
Organizations, Jossey-Bass, 2002.
9 Zenas Block and Ian MacMillan, Corporate Venturing: Creating New
Businesses within the Firm, Harvard Business School Press, 1993,
Chapter 9, Controlling the Venture.
10 Richard Leifer, Christopher McDermott, Gina Colarelli, Lois Peters,
Mark Rice, and Robert Veryzer, Radical Innovation: How Mature
Companies Can Outsmart Upstarts, Harvard Business School Press.
314
INDEX
3M xi, 1, 39, 59, 155, 181, 184–5 Barrett, Craig 10–11, 14–15
BAT 62, 180
AA 22 Beefeater 191
Accel-KKR 13 Bell Labs 211
Acer 175–6 Bell, Charlie 16
adjacencies steps test 60–61 Berkshire Hathaway 264
Advanced Digital Devices 15 BG 5, 21, 52–4, 126–7, 211
advantage, looking for 27–9 BG Corporate Development
Alcatel 207 Division 225
Alexander Howden 47–9, 134 BG Property 211
Alexander, Marcus 1, 156 Birkinshaw, Julian 5, 23, 269
alien territory 161 Block, Zenas 272–4
Allen, James 125, 294–7 Blockbuster 126
Allen, Paul 18 Boeing 153
AlliedSignal 32 Boots 96, 147–8, 163, 166–7, 168
Alza 138 Boots Healthcare International
ambidextrous organization 232 168
ambitions, realism about 33–4 Borjesson, Rolf 138
Anglo American 203 Bossidy, Larry 32
Ansoff, Igor 154, 159, 161–2 Boston Box 152–3
Apple xi, 127 Boston Consulting Group 4, 152
Aroma Café 12 Boston Market 12, 15
Ashridge Portfolio Display 159–62, BP 24
176 Brandon, Stephen 52–3
Ashridge Strategic Management Branson, Sir Richard x, 181
Centre 3, 35 Brewer’s Fayre 191
ASM Lithography 5 Bristol Myers Squibb 184
AstraZeneca 180 British Airways 20, 29
AT&T 42 British Sugar 55–6, 94–5
attention tests for existing businesses BSO/Origin 5
142–3 BT 211–13
attractiveness of new markets 165 BT Brightstar 211–12, 214
Buffett, Warren 264
Baghai, Merhdad 270–72 Burgelman, Robert 9–10, 14, 182,
Bain & Co. 19, 63, 125 193–4, 196–8, 233–4, 244–5, 269,
ballast businesses 161 275–9
Barclays Bank 2 Burger King 133
INDEX
business model reasons for failure 209–10

potential for high margins role of 227–30
120–22 Corporate-Level Strategy 1–2
vulnerability of 127–8 Courtaulds 24–5
Crown Cork and Seal 25–6, 85–7
Cadbury 96, 110, 113, 124, 126,
128, 139, 147–8 David Lloyd Leisure 46, 191
calibration of profit pool potential Davis, Sir Peter 134
129–30 Dell x, 265
Calvin Klein 246 Dell, Michael xiii
Canon 1, 39–40, 180–81 Diageo 207
Cantalupo, Jim 12, 15–16 Diner Inside 13, 20
Cemex xiv Direct Line 23, 46–50, 134, 175
Centrica 22, 141 Disney ix, x–xi
Chase Manhattan 72 disruptive technologies 169–76, 177,
Cheng Uei Precision xi 219–20
Chipotle Mexican Grill 12, 15, 144 and impact on existing businesses
Christensen, Clayton 42, 65–6, 174–6
169–76, 177, 233–4, 269 and leaders/sponsors 174
Chrysler 145 and profit pool potential 173–4
Churchill 135 and value advantage 172–3
Cisco xiii distraction risks 142–7, 165, 175,
clarity xii 201–2, 203
Coca-Cola 72, 153, 157 diversification 1, 149, 150–77
Coley, Stephen 270–72 Dixons 50, 120, 135, 162, 175
Coller Capital 211–12, 214 dogs 26–7, 94, 117, 120, 129, 165,
Collins, Jim 33 202
commitment xii–xiii Donatos Pizza 12, 15
Compaq 207 DSM 3–4, 68, 162, 202
competitor behavior 27 DuPont 43
Confidence Check xii, 149, 251–62 Dyess, Kirby 27
execution confidence 253–60
Traffic Lights confidence 252–3 Eastman Kodak 17, 31
Connelly, John 85–7 easyJet 29
contribution of competitors 164 eBay xiii
core businesses, investing in 24–6 ecosystem venturing 208, 215–18,
core competencies 59–60, 98–9, 155 228
Corning 39–40 edge-of-heartland businesses 159,
corporate incubators 63, 207 162, 166, 167, 168, 178–9
Corporate Strategy Board 19, 71–2, efficient market hypothesis 67–8
90–91 Egg 4, 32, 134, 145, 235, 242
corporate venturing 5, 6–7, 23–4, Elizabeth Arden 160–61
27, 63–4, 182, 207–30 eMac Digital 13
316
THE GROWTH GAMBLE
Emerging Business Opportunities Grove, Andy 9–11, 197

program 90, 137, 194–5, 198, growth gamble 19–21
248–9 growth restarts 19, 90–91
Emerson 1 growth stalls 71–2, 73
Enron 53, 263 growth
Ericsson 207 arguments for 66–70
Exxon 159, 161 low 6–7
moral arguments for 66–70
Fazoli’s 13, 15 new sources of 2
First Direct 32, 134 value-driven arguments for 66–7,
fit 1, 21, 59–62, 165, 233 70–72
five forces 56–7, 122–4, 173 Guinness 168
Flaherty, Mike 47–50, 134
Fleming 72 Hamel, Gary ix–xv, 63, 135, 155,
focus xiii–xv, 1 185, 265, 269, 282–4
Food.com 13 Hammond, Tim 95–6
Ford ix, 71 Hanson 154
Foster, Richard 43, 279–82 Harris, Mike 32, 134
Freeserve 120, 135, 175 harvest venturing 208, 210–14, 228
Haveland, Roy 47–50, 134
Gamechanger 218–20 heartland businesses 159, 166, 167,
Gates, Bill 18, 38, 264 168, 178–9
GE x, 1, 2, 22–3, 40, 43, 141, 162, Hewlett-Packard 1, 22–3, 30, 94,
167, 168 166, 168
GE Capital 27 Homeserve 132–3
GE Equity 221–4 Honda 59
GEC/Marconi 141 hubris 57–62
General Management School
150–54, 156, 158, 159 IBM 18, 22–3, 31, 34, 50, 88–90, 94,
Gerstner, Louis 31, 88–9, 264 135, 137, 166, 175–6, 194–5,
Gibbons, Barry 133 198, 206, 248–9
Gill, Frank 9–11 IBM Global Services 31
Gillette 2 ICI 71
Go 20, 29 IDV 30
goals, ambitious 33–4 imagination xiii–xv
Goldfish 141 impact on existing businesses 95,
Google xiii 97–8, 140–47, 165, 167, 169
Goold, Michael 1, 156, 234 industry environments matrix
Gorilla Park 27 199–203
GrandMet 30, 140, 167, 168 industry structure, potential for high
Green Giant 140 margins 122–5
Greenberg, Jack 11–12, 15 innnovation xv
Grossman, David 135 types of 171–2
317
INDEX
innovation venturing 208–9, 218–21, 69, 146, 181–2, 226

224, 229 Legal & General 2
Integrated Circuits 5 Leifer, Richard 258–9, 287–9
Intel ix, 2, 6, 8–18, 19–20, 24, 28, leveraging x–xi
33, 40, 41, 42, 45, 51, 71, 72, 80, LG 70
123, 127, 176, 178, 179, 181, life cycle, corporate 45–6
183, 184, 193, 197, 199, 201, Ling, Jimmy 263
207, 226 Lloyds TSB 141, 219
Intel Capital 10, 215–17 Lloyds TSB Strategic Ventures
Intel Data Services 14 219–20
Intel Products Group 9–10 Lotus Corporation 89
Iridium 261, 266 low growth 65–91, 92
Irish Life 140–41 choosing 83–90
ITT 151–2 financial consequences of 73–8
low-growth companies, becoming
Johnson & Johnson 72, 138, 207, high growth 90–91
215–16 LTV 263
Johnson and Johnson Lucent New Ventures Group
Development Corporation 211–13, 214
215–17
Johnson Matthey 184, 187 MacMillan, Ian 272–4
Johnson, John 10 management 130–39
joint ownership 157 importance of 31–3
joint ventures xii Mannesman 1–2, 3, 40, 162
Juran, Joseph 265 Marconi 263
Mario, Ernest 138
Kanter, Rosabeth 194, 284–7 market expectations 78–82
Kaplan, Sarah 43, 279–82 highly rated companies 78–80
Keebler 20, 25, 29 lowly rated companies 81–2
Kelloggs 2, 20, 25, 29 Marriott 46, 191
Kerry Foods 66–7 Mars 61–2, 96, 108–9, 110, 113–15,
Kiewit, Peter 87 124, 126, 128, 139, 146, 147–8,
Kroc, Ray 11 159, 240, 242, 243
Mason, Heidi 269
Laurie, Don 269 Matsushita 123
leader, opportunity to be McCafe 13, 20
125–6 McCaw Cellular 42
leadership quality 131–6 McDermott, Christopher 287–9
leadership/sponsorship quality 94–5, McDonald’s 2, 6, 8–18, 19–20, 21,
97–8, 130–39, 165, 167, 169, 203 24, 28, 33, 38, 40, 41, 45, 51, 54,
learning costs 29–31, 110–17, 164, 59, 69, 72, 80, 144–6, 153, 157,
167, 169, 173 176, 178, 181, 183, 184, 199,
Lederhausen, Mats 13, 15–16, 38, 201–2, 208, 226, 269–97
318
THE GROWTH GAMBLE
McDonald’s International 12 importance to strategy 242

McDonald’s Ventures 15–16, integration with existing busi-
146, 208, 226 nesses 232–40
McManus, Charles 85 layers between investor and
MediaOne 42 entrepreneur 243–5
Mercedes 145 monitoring progress 250–62
Merck 72 objectivity of reporting relation-
Microsoft ix, 17–18, 71, 76–8, 123, ships 245–6
216, 264 parenting 231–62
Miner, John 14 reporting level 240–46
modeling xii searching for 178–206
Moore, Gordon 9, 197 six rules for 24–34
Morrison, Andy 5 strategy for 41–2
Motorola ix, 71, 261, 266 support from layers above 246–50
Mustang Ventures 217 when to end 260–62
new ideas
Nasser, Jacques ix bottom-up processes for 192–5
National Westminster 2–3 bringing together top-down and
Navteq 5, 261 bottom-up processes 196–8
NCR 42 improving natural flow of 188–98
Neiman Marcus 96, 147–8 managing flow of 204–5
Nemesis 57–62 natural flow of 180–98
Nestlé 71, 108–9, 124, 126, 128, 130 top-down processes for 188–92
net non-tradable contribution 107–9 new leg venturing 209, 213, 218,
new business development process 224–6, 227
28 new-to-the-world opportunities 169
new business failure, avoidable Nokia 2, 207, 221
causes of 50–64 Nokia Early Stage Technology
New Businesses Traffic Lights xi, unit 230
xiv, 35–6, 92–149, 172–6, 183–4, Nokia New Growth Business
186–7, 196–8, 200–203, 205, 251, unit 219–20
259, 264, 267 Nokia Venture Partners 221, 230
new businesses Nokia Ventures Organization
challenge of 8–18 227, 230
creating ix–xv, 39–40 numbers game 62–4
definition of 20 Nutrigrain 29
difficulties of building 42–64
ensuring right processes for 204–6 O’Connor, Gina Colarelli 287–9
existence of skilled managers O’Reilly, Charles 232, 291–4
242–3 odds, beating 19–42
gambling on 41–2 opportunities that fit, shortage of
high failure rate 19–21 20–21, 24, 41
how much to invest in 198–203 opportunities, identifying 31–3
319
INDEX
organization design Pret A Manger xii, 13, 15, 144

accountability test 236, 238, 241 PricewaterhouseCoopers 89
difficult links test 236, 238, 241 private equity venturing 209, 218,
feasibility test 236, 237, 241 221–4, 229
flexibility test 236, 238, 241 Procter & Gamble 3, 43, 71, 155
market advantage test 236, 237, product/market matrix 59–60,
240 154–5, 159, 161–2
nine tests for 235–40 profit pool potential 97–8, 117–30,
parenting advantage test 236, 164–5, 167, 169, 202
237, 240 Prudential 4, 32, 40, 134, 145, 186,
people test 236, 237, 240 235, 242
redundant hierarchy test 236, purposeful accidents 37–8
238, 241
specialist cultures test 236, 237, rare games 26–7, 117, 119–20, 129,
240 165, 166, 168–9, 178–9
Otellini, Paul 14–15 relatedness 154–5, 157, 233
overvaluation trap 78–80, 83–4 Rentokil 28–9
Rentokil Initial 66
Palmisano, Sam 90, 194 research project 3–6, 21–4
parenting new businesses 205–6 resource-based view of strategy 155
Parenting Opportunity Analysis Rexam 138
246–50, 262 Rice, Mark 287–9
Parenting Theory 1, 3–4, 64, Rio Tinto 1, 156
156–66, 176 risks, inherent in new businesses
link with Traffic Lights 163–69 46–50
Parker, Gerry 14 Royal Academy of Arts xv
Partner Brands 13, 15–16, 144–6 Royal Bank of Scotland xi, 2, 23,
Patrick, John 135 46–50, 134, 175
PC Plus 14, 80 Royal Sun Alliance 207
PC World 50 Rumelt, Richard 154
Pearson, Michael 219 Ryanair 29
Peters, Lois 287–9
Philips 2, 5–6, 17, 70, 123, 141, Samsung 265
180–81, 199, 206, 214, 261, 266 saplings 30–31, 166–8, 178–9
Pizza Hut 46, 191 Savelsberg, Rene 214
PLM 138 Schreuder, Hein 3, 68
Polygram 5 Sheffey, Forrest 187
PortalPlayer xi Shell 2, 21, 59, 115, 218–19
Porter, Michael 56–7, 98–9, 122–4 Siemens 141, 217
positive thinking 39 skills, humility about 29–31
potential, releasing 37 Slater Walker 151, 263
potential/fit matrix 95–6 Slater, Jim 263
Prahalad, CK 155 Slywotzky, Adrian 289–91
320
THE GROWTH GAMBLE
sponsorship quality 136–8 tradability 102–5

Spring xi value-based management 154
Steen, John 24 venture capitalists xiii, 20, 40, 63,
strategic business case 35–6 243
Strategos 63, 185 venturing
synergies 140–42, 165, 174–5 anyone can do this syndrome 223
Synergy School 151, 154–6, 156–7, culture change pitfall 220
158, 159 loss of focus temptation 216
new legs pitfall 213
TCI 42 Veryzer, Robert 287–9
Tennant, Anthony 30 Virgin Group x, xi, 39–40, 181,
tentativen ess xii–xiii 184–5
Tesco x, xi Vivendi 53, 263
TGI Friday’s 191 Vodafone 207
The Link 175 Volkholz, Klaus 5–6, 17
Thomson, Sir Clive 66
tradability 163 Wal-Mart 126
Trahar, Tony 203 Walters, Sir Peter 24–5
Transco 104–5 Watson, Thomas Jr. 18
Travel Inn 191 Whitbread 45–6, 95–6, 191–2, 197
Treacy, Michael 269 White, David 270–72
Trumans 30 WHSmith 113
Tushman, Michael 232, 291–4 Williams 104
Wi-Max 14
Uhlman, Tom 211 WL Gore x
undervaluation trap 81–2, 84 Wood, Peter 23, 47–50, 134, 175
Unilever 108, 124, 126, 128, 130,
159, 160–61, 179, 208, 246 Xerox 115
unique contribution 163–4
UPS x Yahoo ix, xiii
Yun-Farmbrough, Simon 4, 145
value added, potential for 159–60
Value Advantage Equation 99–117, Zook, Chris 19, 25, 125, 294–7
172
value advantage 94, 97–117, 163,
167, 169, 202
value creation 83
value destruction 42–4, 50–64, 83,
159–60
value traps 159
value
competitors’ unique 106–8
sources of unique 99–102
321
ACKNOWLEDGMENTS
I
n putting this book together there are many people who deserve
our thanks. Most important are the large number of managers
who gave us their time and who trusted us to treat their confi-
dences with respect.
Julian Birkinshaw, Robert Burgelman, and Gary Hamel have been
particularly useful discussants. Chris Tchen shared his consulting expe-
rience. Lucy Campbell did some excellent research on McDonald’s.
Angela Munro helped assemble the text and exhibits. Michael Goold
provided editorial guidance.
Nicholas Brealey, our publisher, insisted on some structural changes
that have made the book much more accessible. He also worked tire-
lessly on the title and provided the sort of relationship every author
would most want with his or her publisher.

The Growth Gamble

Uploaded by

Copyright:

Available Formats

The Growth Gamble

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Growth Gamble

Uploaded by

Copyright:

Available Formats

FOR SALE & EXCHANGE

Subscribe for FREE download 5000+ trading books.

"Although I do not agree with all of Campbell and Park's views,

"Campbell and Park challenge some well documented views about

"Campbell and Park address a critical issue—new growth for the

Andrew Campbell & Robert Park

3–5 Spafield Street 100 City Hall Plaza, Suite 501

© Andrew Campbell & Robert Park 2005

British Library Cataloguing in Publication Data

All rights reserved. No part of this publication may be reproduced,

Printed in Finland by WS Bookwell.

Foreword by Professor Gary Hamel ix

Chapter 1 The Challenge of New Businesses 8

Chapter 2 Beating the Odds 19

Chapter 3 The Difficulties of Building New Legs 43

Chapter 4 When Low Growth Is Better than Gambling 65

Chapter 5 The New Businesses Traffic Lights 92

Chapter 6 Diversification 150

Chapter 7 Searching for New Businesses 178

Chapter 8 Is There a Role for Corporate Venturing? 207

Chapter 9 Positioning and Supporting a New Business 231

Chapter 10 An Age of Realism 263

Appendix A Advice from Other Authors 269

TAKING THE GAMBLE OUT

companies that have succeeded spectacularly in growing new busi-

IMPROVING THE ODDS

LEVERAGE SOMETHING YOU KNOW HOW TO DO

pany’s deep experience in producing live entertainment for its amuse-

DON’T BE AFRAID TO PARTNER

BE BOTH COMMITTED AND TENTATIVE

Third, be committed and tentative. It’s difficult to get a new business

❏ being extraordinarily clear about what you do and do not know. It is

EXPAND YOUR IMAGINATION AND NARROW YOUR

in the rapidly growing profusion of uncharted websites. It requires

capable of inventing novel business concepts with the power to posi-

examples of companies diversifying into very different businesses. It

THE RESEARCH BEGINNINGS

The research into DSM’s transitions was fascinating. It reinforced

Three other people have been particularly influential in this project.

inappropriate risks, and overlook the costs of distracting attention

HOW TO READ THE BOOK

This book is for a mix of audiences: senior managers, business develop-

choice. This understanding will help managers be wiser when they

THE INTEL STORY

included new businesses such as motherboards for PCs, LAN adapter

Despite the emergence of many new opportunities across the comput-

Overall, Andy Grove concluded that he had failed in his ambition to

THE MCDONALD’S STORY

McDonald’s has had around 40 years of exceptional performance.

However, he was worried about the company’s ability to continue its

We owe it to our shareholders, owner/operators, employees and sup-

In 2000 and 2001 a number of other initiatives followed:

❏ The acquisition, from Chapter 11 bankruptcy, of Boston Market, a

❏ An investment in Food.com, a company known for its internet food

This rash of initiatives was partly boosted by a decision in 1999 to give

The pipeline of new businesses must be dynamic and vibrant. There

RECENT HISTORY AT INTEL AND MCDONALD’S

In early 1999, Craig Barrett decided to focus Intel’s commitment to

television business because of delays in its next-generation computer

Lederhausen, managing director of McDonald’s Ventures, has been