59 Customizing global marketing "The big issue today is not whether to go global but how to tailor the global marketing concept to fit each business." John A. Quelch and Edward /. Hoff In the best of all possible worlds, mar- keters would only have to come up with a great product and a convincing marketing program and they would have a worldwide winner. But despite the obvious economies and efficiencies they could gain with a stan- dard product and program, many managers fear that global marketing, as popularly defined, is too extreme to be practical. Because customers and competitive conditions differ across countries or because powerful local managers will not stand for centralized decision making, they argue, global marketing just won't work. Of course, global marketing has its pit- falls, but it can also yield impressive advantages. Stan- dardizing products can lower operating costs. Even more important, effective coordination can exploit a company's best product and marketing ideas. Too often, executives view global mar- keting as an either/or proposition-either full standard- ization or local control. But when a global approach can fall anywhere on a spectrum from tight worldwide coordination on programming details to loose agree- ment on a product idea, why the extreme view? In ap- plying the global marketing concept and making it work, flexibility is essential. Managers need to tailor the approach they use to each element of the business system and marketing program. For example, a manu- facturer might market the same product under differ- ent brand names in different countries or market the same brands using different product formulas. Mr. Quelch is an associate professor of business administration at the Harvard Business School where he teaches in the new Multinationa} Marketing Management executive program. This is his sixth HBR arti- cle, the last being "How to Build a Product Licensing Pro- gram" (May-June 1985). Mr. Hoff is a PhD candidate in business economics at Harvard University. He was an instructor in marketing at the Harvard Business School, which awarded him a Dean's Doctoral Fellowship to complete his PhD. The big issue today is not whether to go global but bow to tailor the global marketing concept to fit each business and how to make it work. In this article, we'll first provide a framework to help manag- ers think about how they should structure the different areas of the marketing function as the business shifts to a global approach. We will then show how compa- nies we have studied are tackling the implementation challenges of global marketing. How far to go How far a company can move toward global marketing depends a lot on its evolution and tra- ditions. Consider these two examples: n Although the Coca-Cola Company had conducted some international business before 1940, it gained true global recognition during World War 11, as Coke bottling plants followed the march of U.S. troops around the world. Ma.

1. 59
Customizing
global
marketing
"The big issue today is
not whether to go global but how
to tailor the global
marketing concept to
fit each business."
John A. Quelch and
Edward /. Hoff
In the best of all possible worlds, mar-
keters would only have to come up with a great product
and a convincing marketing program and they would
have a worldwide winner. But despite the obvious
economies and efficiencies they could gain with a stan-
dard product and program, many managers fear that
global marketing, as popularly defined, is too extreme
to be practical. Because customers and competitive
conditions differ across countries or because powerful
local managers will not stand for centralized decision
making, they argue, global marketing just won't work.
Of course, global marketing has its pit-
falls, but it can also yield impressive advantages. Stan-
dardizing products can lower operating costs. Even
more important, effective coordination can exploit a

2. company's best product and marketing ideas.
Too often, executives view global mar-
keting as an either/or proposition-either full standard-
ization or local control. But when a global approach
can fall anywhere on a spectrum from tight worldwide
coordination on programming details to loose agree-
ment on a product idea, why the extreme view? In ap-
plying the global marketing concept and making it
work, flexibility is essential. Managers need to tailor
the approach they use to each element of the business
system and marketing program. For example, a manu-
facturer might market the same product under differ-
ent brand names in different countries or market the
same brands using different product formulas.
Mr. Quelch is an associate professor of
business administration at the Harvard Business School
where he teaches in the new Multinationa} Marketing
Management executive program. This is his sixth HBR arti-
cle, the last being "How to Build a Product Licensing Pro-
gram" (May-June 1985).
Mr. Hoff is a PhD candidate in business
economics at Harvard University. He was an instructor in
marketing at the Harvard Business School, which awarded
him a Dean's Doctoral Fellowship to complete his PhD.
The big issue today is not whether to go
global but bow to tailor the global marketing concept
to fit each business and how to make it work. In this
article, we'll first provide a framework to help manag-
ers think about how they should structure the different
areas of the marketing function as the business shifts
to a global approach. We will then show how compa-
nies we have studied are tackling the implementation

3. challenges of global marketing.
How far to go
How far a company can move toward
global marketing depends a lot on its evolution and tra-
ditions. Consider these two examples:
n Although the Coca-Cola Company had
conducted some international business before 1940, it
gained true global recognition during World War 11, as
Coke bottling plants followed the march of U.S. troops
around the world. Management in Atlanta made all
strategic decisions then-and still does now, as Coca-
Cola applies global marketing principles, for example,
to the worldwide introduction of Diet Coke. The hrand
name, concentrate formula, positioning, and advertis-
ing theme are virtually standard worldwide, but the ar-
tificial sweetener and packaging differ across countries.
Local managers are responsible for sales and distribu-
tion programs, which they run in conjunction with lo-
cal bottlers.
n The Nestle approach also has its roots
in history. To avoid distribution disruptions caused by
wars in Europe, to ease rapid worldwide expansion, and
to respond to local consumer needs. Nestle granted its
local managers considerable autonomy from the out-
Harvard Business Review May-Iunc 1986
set. Wbile the local managers still retain mucb of that
decision-making power today, Nestle headquarters at
Vevey has grown in importance. Nestle has transferred

4. to its central marketing staff many former local manag-
ers who had succeeded in their local Nestle husinesses
and who now influence country executives to accept
standard new product and marketing ideas. The trend
seems to be toward tighter marketing coordination.
To conclude that Coca-Cola is a global
marketer and Nestle is not would be simplistic. In Ex-
hibit I, we assess program adaptation or standardiza-
tion levels for each company's business functions,
products, marketing mix elements, and countries. Each
company has tailored its individual approach. Further-
more, as Exhibit I can't show, the situations aren't
static. Readers can themselves evaluate their own cur-
rent and desired levels of program adaptation or stan-
dardization on these four dimensions. The gap between
the two levels is the implementation challenge. The
size of the gap-and the urgency with which it must be
closed-will depend on a company's strategy and finan-
cial performance, competitive pressures, technological
change, and converging consumer values.
Four dimensions of global
marketing
Now let's look at the issues that arise
when executives consider the four dimensions shown
in Exhibit I in light of the degree of standardization or
adaptation that is appropriate.
Business (unctions. A company's ap-
proach to global marketing depends, first, on its overall
business strategy. In many multinationals, some func-
tional areas have greater program standardization than
others. Headquarters often controls manufacturing, fi-
nance, and RikD, while the local managers make the

5. marketing decisions. Marketing is usually one of the
last tunctions to be centrally directed. Partly because
product quality and accounting data are easier to mea-
sure than marketing effectiveness, standardization can
be greater in production and finance.
Products. Products that enjoy high scale
economies or efficiencies and are not highly culture-
bound are easier to market globally than others.
1 Economies or efficiencies. Manufactur-
ing and R&D scale economies can result in a price
spread between the global and the local product that is
too great for even the most culture-bound consumer to
resist. In addition, management often has neither the
time nor the Ri^D resources to adapt products to each
country. The markets for high-tech products like com-
puters are not only very competitive but also affected
by rapid technological change.
Most packaged consumer goods are less
susceptible than durable goods like televisions and
cars to manufacturing or even R&D economies. Coca-
Cola's global policy and Nestle's interest in tighter
marketing coordination are driven largely by a desire
to capitalize on the marketing ideas their managers
around the world generate rather than by potential
scale economies. Nestle, for example, manufactures its
packaged soups in dozens of locally managed plants
around the world, with some transference of engineer-
ing know-how through a headquarters staff. Products
and marketing programs are also locally managed, but
new ideas are aggressively transferred, with local man-
agers encouraged-or even prodded-to adapt and use
them in their own markets. For Nestle, global market-

6. ing does not so much yield high manufacturing econo-
mies as high efficiency in using scarce new ideas.
2 Cultural grounding. Consumer products
used in the h o m e - like Nestle's soups and frozen foods
- a r e often more culture-bound than products used
outside the home such as automobiles and credit cards,
and industrial products are inherently less culture-
bound than consumer products. (Products like personal
computers, for example, are often marketed on the ba-
sis of performance benefits that share a common tech-
nical language worldwide.) Experience also suggests
that products will be less culture-bound if they are
used by young people whose cultural norms are not
ingrained, people who travel in different countries, and
ego-driven consumers who can be appealed to through
myths and fantasies shared across cultures.
Exhibit I lists four combinations of the
scale economy and cultural grounding variables in or-
der of their susceptibility to global marketing. Manag-
ers shouldn't be bound by any matrix, however; they
should find creative ways to prepare a product for glob-
al marketing. If a manufacturer develops a new version
of a seemingly culture-bound product that is based on
new capital-intensive technology and generates supe-
rior performance benefits, it may well be possible to
introduce it on a standard basis worldwide. Procter &
Gamble developed Pampers disposable diapers as a
global brand in a product category that intuition would
say was culture-bound.
Marketing mix elements. Few consumer
goods companies go so far as to market the same prod-
ucts using the same marketing program worldwide.
And those that do, like Lego, the Danish manufacturer

7. of construction toys, often distribute their products
through sales companies rather than full-fledged mar-
keting subsidiaries.
Global marketing 61
Exhibit I Global marketing planning matrix:
how far to go
Adaptation
Full Partial
Standardization
Partiai Full
Business
functions
Products
Research and
development
Finance and
accounting
Manufacturing
Procurement
Marketing

8. Low cultural
grounding
High
economies or
efficiencies
Low cultural
grounding
Low
economies or
efficiencies
High cultural
grounding
High
economies or
efficiencies
High cultural
grounding
Low
economies or
efficiencies
Marketing mix
elements
Product
design
Countries
Region 1
Region 2
Nestle Coca-Cola

9. 62 Harvard Business Review May-]une 1986
For most products, the appropriate de-
gree of standardization varies from one element of the
marketing mix to another. Strategic elements like
product positioning are more easily standardized than
execution-sensitive elements like sales promotion. In
addition, when headquarters believes it has identified a
superior marketing idea, whether it be a package de-
sign, a brand name, or an advertising copy concept, the
pressure to standardize increases.
Marketing can usually contribute to
scale economies most significantly by creating a stan-
dard product design that will sell worldwide, permit-
ting savings through globalized production. In addi-
tion, scale economies in marketing programming can
be achieved through standard commercial executions
and copy concepts. McCann-Erickson claims to have
saved $90 million in production costs over 20 years by
producing worldwide Coca-Cola commercials. To en-
sure that they have enough attention-getting power to
overcome their foreign origins, however, marketers of-
ten have to make worldwide commercials expensive
productions.
To compensate local management for
having to accept a standard product and to fit the core
product to each local market, some companies allow
local managers to adapt those marketing mix elements
that aren't subject to significant scale economies. On
the other hand, local managers are more likely to ac-
cept a standard concept for those elements of the mar-

10. keting mix that are less important and, ironically, often
not susceptible to scale economies. Overall, then, the
driving factor in moving toward global marketing
should be the efficient worldwide use of good market-
ing ideas rather than any scale economies from stan-
dardization.
In judging how far to go in standardizing
elements of the marketing mix, managers must also be
mindful of the interactions among them. For example,
when a product with the same brand name is sold in
different countries, it can be difficult and sometimes
impossible to sell them at different prices.
Countries. How far a decentralized
multinational wishes to pursue global marketing will
often vary from one country to another. Naturally,
headquarters is likely to become more involved in
marketing decisions in countries where performance is
poor. But performance aside, small markets depend
more on headquarters assistance than large markets.
Because a standard marketing program is superior in
quality to what local executives, even with the benefit
of local market knowledge, could develop themselves,
they may welcome it.
Large markets with strong local man-
agements are less willing to accept global programs.
Yet these are the markets that often account for most
of the company's investment. To secure their accep-
tance, headquarters should make standard marketing
programs reflect the needs of large rather than small
markets. Small markets, being more tolerant of devia-
tions from what would be locally appropriate, are less
likely to resist a standard program.

11. As we've seen, Coca-Cola takes the same
approach in all markets. Nestle varies its approach in
different countries depending on the strength of its
market presence and each country's need for assis-
tance. In completing the Exhibit I planning matrix,
management may decide that it can sensibly group
countries by region or by stage of market development.
Too far too fast
Once managers have decided how glob-
al they want their marketing program to be, they
must make the transition. Debates over the size of the
gap between present and desired positions and the
speed with which it must be closed will often pit the
field against headquarters. Such conflict is most likely
to arise in companies where the reason for change is
not apparent or the country managers have had a lot of
autonomy. Casualties can occur on both sides:
n Because Black &. Decker dominated the
European consumer power tool market, many of the
company's European managers could not see that a
more centrally directed global marketing approach was
needed as a defense against imminent Japanese compe-
tition. To make his point, the CEO had to replace sev-
eral key European executives.
n In 1982, the Parker Pen Company, forced
by competition and a weakening financial position to
lower costs, more than halved its number of plants and
pen styles worldwide. Parker's overseas subsidiary
managers accepted these changes but, when pressed to
implement standardized advertising and packaging,
they dug in their heels. In 1985, Parker ended its much

12. heralded global marketing campaign. Several senior
headquarters managers left the company.
If management is not careful, moving
too far too fast toward global marketing can trigger
painful consequences. First, subsidiary managers who
joined the company because of its apparent commit-
ment to local autonomy and to adapting its products
to the local environment may become disenchanted.
When poorly implemented, global marketing can
make the local country manager's job less strategic.
Second, disenchantment may reinforce not-invented-
here attitudes that lead to game playing. For instance,
some local managers may try bargaining witb head-
Global marketing 63
quarters, trading the speed with which they will accept
and implement the standard programs for additional
budget assistance. In addition, local managers compet-
ing for resources and autonomy may devote too much
attention to second-guessing headquarters' "hot but-
tons." Eventually the good managers may leave, and
less competent people who lack the initiative of their
predecessors may replace them.
A vicious circle can develop. Feeling
compelled to review local performance more closely,
headquarters may tighten its controls and reduce re-
sources without adjusting its expectations of local
managers. Meanwhile, local managers trying to gain
approval of applications for deviations from standard
marketing programs are being frustrated. The expand-
ing headquarters hureaucracy and associated overhead

13. costs reduce the speed with which the locals can re-
spond to local opportunities and competitive actions.
Slow response time is an especially serious problem
with products for which barriers to entry for local com-
petitors are low.
In this kind of system, weak, insecure
local managers can become dependent on headquarters
for operational assistance. They'll want headquarters
to assume the financial risks for new product launches
and welcome the prepackaged marketing programs. If
performance falls short of headquarters' expectations,
the local management can always blame the failure on
the quality of operational assistance or on the standard
marketing program. The local manager who has clear
autonomy and profit-and-loss responsibility cannot
hide behind such excuses.
If headquarters or regions assume much
of the strategic burden, managers in overseas subsidi-
aries may think only about short-term sales. This focus
will diminish their ability to monitor and communi-
cate to headquarters any changes in local competitors'
strategic directions. When their responsibilities shift
from strategy to execution, their ideas will become less
exciting. If the field has traditionally been as important
a source of new product ideas as the central R6LD lab-
oratory, the company may find itself short of the grass-
roots creative thinking and marketing research infor-
mation that R&X) needs. The fruitful dialogue that
characterizes a relationship between equal partners
will no longer flourish.
How to get there
when thinking about closing the gap

14. between present and desired positions, most execu-
tives of decentralized multinationals want to accom-
modate their current organizational structures. They
rightly view their subsidiaries and the managers who
run them as important competitive strengths. They
generally do not wish to transform these organizations
into mere sales and distribution agencies.
How then in moving toward global mar-
keting can headquarters build rather than jeopardize
relationships, stimulate rather than demoralize local
managers? The answer is to focus on means as much as
ends, to examine the relationship between the home
office and the field, and to ask what level of headquar-
ters intervention for each business function, product,
marketing mix element, and country is necessary to
close the gap in each.
As Exhibit U indicates, headquarters
can intervene at five points, ranging from informing to
directing. The five intervention levels are cumulative;
for headquarters to direct, it must also inform, per-
suade, coordinate, and approve. Exhibit U shows the
approaches Atlanta and Vevey have taken. Moving
from left to right on Exhibit 11, the reader can see that
things are done increasingly by fiat rather than patient
persuasion, through discipline rather than education.
At the far right, local subsidiaries can't choose whether
to opt in or out of a marketing program, and headquar-
ters views its country managers as subordinates rather
than customers.
When the local managers tightly con-
trol marketing efforts, multinational managers face
three critical issues. In the sections that follow, we'll

15. take a look at how decentralized multinationals are
working to correct the three problems as they move
along the spectrum from informing to directing.
Inconsistent brand identities. If head-
quarters gives country managers total control of their
product lines, it cannot leverage the opportunities that
multinational status gives it. The increasing degree to
which consumers in one country are exposed to the
company's products in another won't enhance the cor-
porate image or brand development in the consumers'
home country.
Limited product focus. In the decentral-
ized multinational, the field line manager's ambition is
to become a country manager, which means acquiring
multiproduct and multifunction experience. Yet as the
pace of technological innovation increases and the like-
lihood of global competition grows, multinationals
need worldwide product specialists as well as execu-
tives willing to transfer to other countries. Nowhere is
the need for headquarters guidance on innovative orga-
nizational approaches more evident than in the area of
product policy.
Slow new product launches. As global
competition grows, so does the need for rapid world-
wide rollouts of new products. The decentralized
multinational that permits country managers to
proceed at their own pace on new product introduc-
64 Harvard Business Review May-June 1986
Exhibit II Global marketing planning matrix:

16. how to get there
Informing Persuading Coordinating Approving Directing
Business
functions
Products Low cuiturai
grounding
Higfi
economies or
efficiencies
Low cuiturai
grounding
Low
economies or
efficiencies
High cuiturai
grounding
High
economies or
efficiencies
iHigh cuiturai
grounding
Low
economies or
efficiencies
Marketing mix
elements
Product
design

17. Countries
Region 1
Region 2
Country E
Nesti6 Coca-Coia
Global marketing
tions may be at a competitive disadvantage in this new
environment.
Word of mouth
The least threatening, loosest, and
therefore easiest approach to global marketing is for
headquarters to encourage the transfer of information
between it and its country managers. Since good ideas
are often a company's scarcest resource, headquarters
efforts to encourage and reward their generation, dis-
semination, and application in the field will build both
relationships and profits. Here are two examples:
n Nestle publishes quarterly marketing
newsletters that report recent product introductions
and programming innovations. In this way, each sub-
sidiary can learn quickly about and assess the ideas of
others. (The best newsletters are written as if country
organizations were talking to each other rather than as
if headquarters were talking down to the field.)

18. • Johnson Wax holds periodic meetings of
all marketing directors at corporate headquarters twice
a year to build global esprit de corps and to encourage
the sharing of new ideas.
By making the transfer of information
easy, a multinational leverages the ideas of its staff and
spreads organizational values. Headquarters has to be
careful, however, that the information it's passing on is
useful. It may focus on updating local managers about
new products, when what they mainly want is infor-
mation on the most tactical and country-specific ele-
ments of the marketing mix. For example, the concen-
tration of the grocery trade is much higher in the
United Kingdom and Canada than it is in the United
States. In this case, managers in the United States can
leam from British and Canadian country managers
about how to deal with the pressures for extra mer-
chandising support that result when a few powerful re-
tailers control a large percentage of sales. Likewise,
marketers in countries with restrictions on mass me-
dia advertising have developed sophisticated point-of-
purchase merchandising skills that could he useful to
managers in other countries.
By itself, however, information sharing
is often insufficient to help local executives meet the
competitive challenges of global marketing.
Friendly persuasion
Persuasion is a first step managers can
take to deal with the three problems we've outlined.
Any systematic headquarters effort to influence local
managers to apply standardized approaches or

19. introduce new global products while the latter retain
their decision-making authority is a persuasion ap-
proach.
Unilever and CPC International, for ex-
ample, employ world-class advertising and marketing
research staff at headquarters. Not critics but coaches,
these specialists review the subsidiaries' work and try
to upgrade the technical skills of local marketing de-
partments. They frequently visit the field to dissemi-
nate new concepts, frameworks, and techniques, and to
respond to problems that local management raises. (It
helps to build trust if headquarters can send out the
same staff specialists for several years.)
Often, when the headquarters of a de-
centralized multinational identifies or develops a new
product, it has to persuade the country manager in a
so-called prime-mover market to invest in the launch.
A successful launch in the prime-mover market will,
in turn, persuade other country managers to introduce
the product. The prime-mover market is usually se-
lected according to criteria including the commitment
of local management, the probabilities of success, the
credibility with which a success would be regarded hy
managers in other countries, and its perceived trans-
ferahility.
Persuasion, however, has its limitations.
Two problems recur with the prime-mover approach.
First, hy adopting a wait-and-see attitude, country
managers can easily turn down requests to be prime-
mover markets on the grounds of insufficient resources.
Since the country managers in the prime-mover mar-
kets have to risk their resources to launch the new
products, they're likely to tailor the product and mar-

20. keting programs to their own markets rather than to
global markets. Second, if there are more new products
waitmg to be launched than there are prime-mover
markets to launch them, headquarters product special-
ists arc likely to give in to a country manager's de-
mands for local tailoring. But hecause of the need for
readaptation in each case, the tailoring may delay roll-
outs in other markets and allow competitors to pre-
empt the product. In the end, management may sacri-
fice long-term worldwide profits to maximize short-
term profits in a few countries.
Marketing to the same drummer
To overcome the limits of persuasion,
many multinationals are coordinating their marketing
programs, whereby headquarters has a structured role
in both decision making and performance evaluation
that is far more influential than person-to-person per-
suasion. Often using a matrix or team approach, head-
66 Harvard Business Review May-June 1986
quarters shares with country managers the responsi-
bility and authority for programming and personnel
decisions.
Nestle locates product directors as well
as support groups at headquarters. Together they de-
velop long-term strategies for each product category on
a worldwide basis, coordinate worldwide market re-
search, spot new product opportunities, spark the field
launch of new products, advise the field on how head-
quarters will evaluate new product proposals, and

21. spread the word on new products' performance so that
other countries will he motivated to launch them.
Even though the product directors arc staff executives
with no line authority, hecause they have all been suc-
cessful line managers in the field, they have great cred-
ibility and influence.
Country managers who cooperate with
a product director can quickly become heroes if they
successfully implement a new idea. On the other hand,
while a country manager can reject a product director's
advice, headquarters will closely monitor his or her
performance with an alternative program. In addition,
within the product category in which they specialize,
the directors have influence on line management ap-
pointments in the field. Local managers thus have to
he concerned about their relationships with head-
quarters.
Some companies assign promising local
managers to other countries and require would-be local
managers to take a tour of duty at headquarters. But
such personnel transfer programs may run into barri-
ers. First, many capable local nationals may not he in-
terested in working outside their countries of origin.
Second, powerful local managers are often unwilling to
give up their best people to other country assignments.
Third, immigration regulations and foreign service re-
location costs arc burdensome. Fourth, if transferees
from the field have to take a demotion to work at head-
quarters, the costs in ill will often exceed any gains in
cross-fertilization of ideas. If management can resolve
these problems, however, it will find that creating an
international career path is one of the most effective
ways to develop a global perspective in local managers.

22. To enable their regional general manag-
ers to work alongside the worldwide product directors,
several companies have moved them from the field to
the head office. More and more companies require re-
gional managers to reach sales and profit targets for
each product as well as for each country within their
regions. In the field, regional managers often focus on
representing the views of individual countries to head-
quarters, hut at headquarters they become more con-
cerned with ensuring that the country managers are
correctly implementing corporatewide policies.
Recently, Fiat and Philips N.V, among
others, consolidated their worldwide advertising into a
single agency. Their objectives are to make each prod-
uct's advertising more consistent around the world and
to make it easier to transfer ideas and information
among local agency offices, country organizations, and
headquarters. Use of a single agency (especially one
that bills all advertising expenditures worldwide) also
symholizes a commitment to global marketing and
more centralized control. Multinationals shouldn't,
however, use their agencies as Trojan horses for greater
standardization. An undercover operation is likely to
jeopardize agency-client relations at the country level.
While working to achieve global coordi-
nation, some companies are also trying to tighten coor-
dination in particular regions:
n Kodak recently experimented by con-
solidating 17 worldwide product line managers at cor-
porate headquarters. In addition, the company made
marketing directors in some countries responsihle for
a line of business in a region as well as for sales of all

23. Kodak products in their own countries. Despite these
new appointments, country managers still retain
profit-and-loss responsihility for their own markets.
Whether a matrix approach such as this
broadens perspectives rather than increases tension
and confusion depends heavily on the corporation's co-
hesiveness. Such an organizational change can clearly
communicate top management's strategic direction,
but headquarters needs to do a persuasive selling joh to
the field if it is to succeed.
n Procter & Gamble has established so-
called Euro Brand teams that analyze opportunities for
greater product and marketing program standardiza-
tion. Chaired by the brand manager from a "lead coun-
try," each team includes hrand managers from other
European subsidiaries that market the brand, managers
from P&G's European technical center, and one of
P&G's three European division managers, each of
whom is responsihle for a portfolio of brands as well as
for a group of countries. Concerns that the larger sub-
sidiaries would dominate the teams and that decision
making would either he paralyzed or produce "lowest
common denominator" results have proved groundless.
Stamped & approved
By coordinating programs with the field,
headquarters can balance the company's local and glob-
al perspectives. Even a decentralized multinational may
decide, however, that to protect or exploit some corpo-
rate asset, the center of gravity for certain elements of
the marketing program should be at headquarters. In
such cases, management has two options: it can send
clear directives to its local managers or permit them to

24. develop their own programs within specified parame-
Global marketing 67
ters and subject to headquarters approval. With a prop-
erly managed approval process, a multinational can
exert effective control without unduly dampening the
country manager's decision-making responsibility and
creativity.
Procter & Gamble recently developed a
new sanitary napkin, and P&G International desig-
nated certain countries in different geographic regions
as test markets. The product, brand name, positioning,
and package design were standardized globally. P&.G
International did, however, invite local managers to
suggest how the global program could be improved and
how the nonglobal elements of the marketing program
should be adapted in their markets. It approved changes
in several markets. Moreover, local managers developed
valuable ideas on such programming specifics as sam-
pling and couponing techniques that were used in all
other countries, including the United States.
Nestle views its brand names as a major
corporate asset. As a result, it requires all brands sold
in all countries to be registered in the home country of
Switzerland. While the ostensible reason for this re-
quirement is legal protection, the effect is that any
product developed in the field has to be approved by
Vevey. The head office has also developed detailed
guidelines that suggest rather than mandate how brand
names and logos should appear on packaging and in
advertising worldwide (with exceptions subject to its

25. approval). Thus the country manager's control over the
content of advertising is not compromised, and the
company achieves a reasonably consistent presenta-
tion of its names and logos worldwide.
The project team countered that by capi-
talizing on potential scale economies, its pan-European
marketing and manufacturing programs would be supe-
rior to any programs the subsidiaries could develop by
themselves. Furthermore, it maintained, the already
developed pan-European program was available off the
shelf. The European sales manager, who was a project
team member, discovered that the salespeople as well
as tradespeople in the target countries wero much more
enthusiastic about the proposed program than the field
marketing managers. So management devised a special
lure for the managers. The project team offered to sub-
sidize the first-year advertising and promotion expen-
ditures of countries launching Sista. Six countries
agreed. To ensure their commitment now that their
financial risk had been reduced, the sales manager in-
vited each accepting country manager to nominate a
member to the project team to develop the final pro-
gram details.
By 1982, the Sista line was sold in 52
countries using a standard marketing program. The
Sista launch was especially challenging because it in-
volved the extension of a product and program already
developed for a single market. The success of the Sista
launch made Henkel's field managers much more re-
ceptive to global marketing programs for subsequent
new products.
Motivating the field

26. Doing it the headquarters way
Multinationals that direct local manag-
ers' marketing programs usually do so out of a sense of
urgency. The motive may be to ensure either that a
new product is introduced rapidly around the world be-
fore the competition can respond or that every man-
ager fully and faithfully exploits a valuable marketing
idea. Sometimes direction is needed to prove that
global marketing can work. Once management makes
the point, a more participative approach is feasible.
In 1979, one of Henkel's worldwide mar-
keting directors wanted to extend the successful Sista
line of do-it-yourself sealants from Germany to other
European countries where the markets were underde-
veloped and disorganized as had once been the case in
Germany. A European headquarters project team vis-
ited the markets and then developed a standard mar-
keting program. The country managers, however, ob-
jected. Since the market potential in each country was
small, they said, they did not have the time or resources
to launch Sista.
Taking into account the nature of their
products and markets, their organizational structures,
and their cultures and traditions, multinationals have to
decide which approach or combination of approaches,
from informing to directing, will best answer their stra-
tegic objectives. Multinational managers must realize,
however, that local managers are likely to resist any pre-
cipitate move toward increased headquarters direction.
A quick shift could lower their motivation and perfor-
mance.
Any erosion in marketing decision mak-

27. ing associated with global marketing will probably be
less upsetting for country managers who have not risen
through the line marketing function. For example, lohn
Deere's European headquarters has developed advertis-
ing for its European country managers for more than a
decade. The country managers have not objected. Most
are not marketing specialists and do not see advertising
as key to the success of their operations. But for coun-
try managers who view control of marketing decision
making as central to their operational success, the tran-
sition will often be harder. Headquarters needs to give
the field time to adjust to the new decision-making
Harvard Business Review May-June 1986
processes that multiuountry brand teams and other
new organizational structures require. Yet management
must recognize that even with a one- or two-year tran-
sition period, some turnover among field personnel is
inevitable. As one German headquarters executive
commented, "Those managers in the field who can't
adapt to a more global approach will have to leave and
run local breweries."
Here are five suggestions on how to
motivate and retain talented country managers when
making the shift to global marketing:
1 Encourage field managers to generate
ideas. This is especially important when R&D efforts
are centrally directed. Use the best ideas from the
fit-Id in global marketing programs (and give recogni-
tion to the local managers who came up with them}.
Unilever's South African subsidiary developed Impulse

28. body spray, now a global brand. R.J. Reynolds revital-
ized Camel as a global brand after the German subsidi-
ary came up with a successful and transferable posi-
tioning and copy strategy
2 Ensure that the field participates in the
development of the marketing strategies and programs
for global brands. A bottom-up rather than top-down
approach will foster greater commitment and produce
superior program execution at the country level.
As we've seen, when P&G International introduced
its sanitary napkin as a global brand, it permitted local
managers to make some adjustments in areas that
were not seen as core to the program, such as coupon-
ing and sales promotion. More important, it encouraged
them to suggest changes in features of the core global
program.
3 Maintain a product portfolio that in-
cludes, where scale economies permit, local as well as
regional and global brands. While Philip Morris's and
Seagram's country managers and their local advertis-
ing agencies are required to implement standard pro-
grams for each company's global brands, the managers
retain full responsibility for the marketing programs of
their locally distributed brands. Seagram motivates
its country managers to stay interested in the global
hrands by allocating development funds to support lo-
cal marketing efforts on these brands and by circulat-
ing monthly reports that summarize market perfor-
mance data by brand and country.
4 Allow country managers continued con-
trol of their marketing budgets so they can respond to
local consumer needs and counter local competition.
When British Airways headquarters launched its £13

29. million global advertising campaign, it left intact the
£18 million worth of tactical advertising budgets that
country managers used to promote fares, destinations,
and ttiur packages specific to their markets. Because
most of the country managers had exhausted their pre-
vious year's tactical budgets and were anxious for fur-
ther advertising support, they were receptive to the
global campaign even though it was centrally directed.
5 Emphasize the general management re-
sponsibilities of country managers that extend beyond
the marketing function. Country managers who have
risen through the line marketing function often don't
spend enough time on local manufacturing operations,
industrial relations, and government affairs. Global
marketing programs can free them to focus on and de-
velop their skills in these other areas. ^
The universal drink
In the postwar years, as Coca-Cola strove mightily
to consolidate its territorial gains, its efforls were
received with mixed feelings. When limited produc-
tion for civilians got under way in the Philippines.
armed guards had to be assigned to the trucks carl-
ing Coke from bottlers to dealers, to frustrate thirsty
outlaws bent on hijacking it. In the Fiji Islands, on
the other hand. Coca-Cola itself was outlawed, at
the instigation of soft-drink purveyors whose busi-
ness had been ruined by the Coke imported for the
solace of G.l.s during the war. Most of the opposi-
tion to the beverage's tidal sweep, however, was
centered in Europe, being provoked by the beer
and wine interests, or by anti-American political
interests, or by a poweriul blend of oenology and
ideology. Today, brewers m England. Spain, and

30. Sweden are themselves bottling Coke, on the if-
you-can't-lick-'em-join-'em principle.... In Western
Europe, Coca-Cola has had to fight a whole series
of battles, varying according to the terrain, not all of
which have yet been won, though victory seems to
be in sight. Before Coca-Cola got rolling in West
Germany, for instance, it had to go to court to halt
the nagging operations of something called the
Coordination Office for German Beverages, which
was churning out defamatory pamphlets with titles
like "Coca-Cola, Karl Marx, and the Imbecility of the
Masses" and the more succinct "Coca-Cola? No!"
In Denmark, lobbyists for the brewers chivied the
Parliament into taxing cola-containing beverages
so heavily that it would have been economically
absurd to try to market Coke there.... At last word,
the Danes were about to relent, though. But in Bel-
gium the caps on bottles of Coke, including bottles
sold at the Brussels Fair, have had to carry, in letters
bigger than those used for Coca-Cola," the forbid-
ding legend "Contient de la cafeine."
From
THE BIG ORINK (Random House)
«n959byE.J.Kahn.Jr
Originally in The New Yorker
Harvard Business Review Notice of Use Restrictions, May 2009
Harvard Business Review and Harvard Business Publishing
Newsletter content on EBSCOhost is licensed for
the private individual use of authorized EBSCOhost users. It is

31. not intended for use as assigned course material
in academic institutions nor as corporate learning or training
materials in businesses. Academic licensees may
not use this content in electronic reserves, electronic course
packs, persistent linking from syllabi or by any
other means of incorporating the content into course resources.
Business licensees may not host this content on
learning management systems or use persistent linking or other
means to incorporate the content into learning
management systems. Harvard Business Publishing will be
pleased to grant permission to make this content
available through such means. For rates and permission, contact
[email protected]
ALL
STRATEGY
IS
True competitive advantages are harder to find and maintain
than people realize.The odds are best in tightly drawn
markets, not big, sprawling ones.
by Bruce Greenwald and Judd Kahn

32. STRATEGIC" IS THE MOST OVERUSED WORD in
thevocabulary of business. Frequently it's just anotherway of
saying, "This is important." The reality is that
there are only a few situations in which companies' strat-
egies affect outcomes. Such situations are, however, worth
trying to create since the alternative, achieving superior
efficiency, is a more demanding route to success, and a
more impermanent one.
The aim of true strategy is to master a market environ-
ment by understanding and anticipating the actions of
other economic agents, especially competitors. But this is
possible only if they are limited in number. A firm that
has privileged access to customers or suppliers or that ben-
efits from some other competitive advantage will have
few of these agents to contend with. Potential competi-
tors without an advantage, ifthey have their wits about
them, will choose to stay away. Thus, competitive advan-
tages are actually barriers to entry. Indeed, the two are, for
all intents and purposes, indistinguishable.
Firms operating in markets without barriers - that is,
where competitive advantages do not exist or cannot be
established - have no choice hut to forget about strategy
and run their businesses as efficiently as possible. Even so,
many neglect operations and divert attention and re-
sources to purportedly strategic moves like acquiring com-
panies in related businesses or entering bigger markets.
In markets without barriers, competition is intense. If
the incumbents have even brief success in earning more
than normal returns on investments, they will find new
entrants swarming in to grab a share ofthe profits. Sooner
or later, the additional competition will push returns
down to the firms' cost of capital. The process that drives
down profits also makes strategy irrelevant since there

33. will be too many other players to take into account and
their roster will always be changing. (See the sidebar "Ef-
ficiency in Place of Strategy.")
Even for companies operating behind solid barriers to
entry, life is not necessarily serene. Ifthe incumbents are
SEPTEMBER 2005 95
All Strategy Is Local
advantages, customer captivity and economies of scale -
which pack an even bigger punch when combined-are
more achievable and sustainable in markets that are re-
stricted in these ways.
Indeed, it's perilous to cbase growth across borders. Be-
cause a global market's dimensions are wider and less de-
fined than a nation's or a region's, firms face a higher risk
of frittering away the advantages they have secured on
smaller playing fields. If a company wants to grow and
still maintain superior returns, the appropriate strategy is
to assemble and dominate a series of discrete but prefer-
ably contiguous markets and then expand only at their
edges. As we wiil show, Wal-Mart's diminishing margins
over the past 15 or so years are strong evidence of the dan-
ger of proceeding otherwise.
The Varieties of
Competitive Advantage
A competitive advantage is something a firm can do that
rivals cannot match. It either generates higher demand or
leads to lower costs. "Demand" competitive advantages
give firms unequaled access to customers. Also known as

34. customer captivity, this type of advantage generally arises
from customers' habits, searching costs, or switching costs.
"Cost" (or "supply") advantages, by contrast, almost al-
ways come down to a superior technology that competi-
tors cannot duplicate-because it is protected by a patent,
for example-or a much larger scale of operation, accom-
panied by declining marginal costs, that competitors can-
not match.
These three factors (customer captivity, proprietary
technology, and economies of scale) generate most com-
petitive advantages. The few other sources-government
support or protection, for instance, and superior access to
information-tend to be limited to particular industries.
Intel benefits from all three fundamental factors. Its
customers, the PC manufacturers, are reluctant to switch
to another supplier because of their long-established re-
lationships with Intel as well as tfieir customers' prefer-
ence, thanks in part to the "Intel Inside" campaign. Intel's
many patents and years of production experience allow
the company to reach a higher yield rate-fewer defects-
in chip production more quickly than its competitors.
And because it can spread the fixed costs of R&D for each
new generation of chips over many more units than its
rivals, it enjoys major economies of scale.
Technological advantages have their limitations,
though. The technologies on which they rest may
rapidly become obsolete. And in cases where such tech-
Bruce Greenwald is the Robert Heilbrunn Professor of Asset
Management and Finance and the director ofthe Heilhrunn
Center for Graham & Dodd Investing at Columbia Business
School in New York. ludd Kahn is a principal of Hummingbird
Management, also in New York. They are the authors

35. 0/Competition Demystified (Portfolio, 2005), from which this
article
is adapted.
well matched, they may try to gain market share by cut-
ting prices, improving services, or making some other
costly move. However, chances are good that they will
succeed only in lowering their returns. Still, such com-
petitors might recognize that the market is roomy
enough not to require head-to-head confrontation at
every turn. Avoiding competition tbat leaves every par-
ticipant worse off is an especially enlightened choice, and
one that deserves to be called "strategic."
The erosion of profitability due to increased competi-
tion from new entrants isn't confined to commodity mar-
kets, as one might expect. It occurs as well in markets for
differentiated products, so long as all actual and potential
competitors have equal access to customers, technology,
and resources. Consider the luxury car market in the
United States. When Cadillac and Lincoln were the only
significant competitors, their brands commanded higher
prices, relative to costs, leading to high returns on in-
vested resources. These returns attracted other competi-
tors to the market: First the Europeans (Jaguar, Mercedes-
Benz, BMW), and then the Japanese (Acura, Lexus, Infitiiti),
started to sell cars in America.
The arrival of these competing products did not lower
prices as it might have for a commodity like copper. Dif-
ferentiation protected against that possibility But prof-
itability still suffered. Cadillac and Lincoln lost sales to the
newcomers. As sales volumes fell, fixed costs per car sold-
such as advettising, product development, special service
support, market Intelligence, and planning-inevitably in-
creased, since these costs had to be covered by the reve-

36. nues from the smaller number of units sold. Margins
fell-same old prices, higher unit costs-so profits took the
double hit of lower margins and reduced sales. If there
were very low barriers to entry, entrants attracted by the
reduced but still above-average return on investment
would have continued to arrive until all the excess profits
were eliminated.
Barriers to entry are easier to maintain in sharply cir-
cumscribed markets. Only within such confines can one
or several firms hope to dominate their rivals and earn su-
perior returns on their invested capital. When competi-
tion is global in scope, the need to circumscribe the com-
petitive arena is even greater. That is why Jack Welch,
instead of just setting revenue and growth targets, insisted
that the only markets in which GE would do business
were ones where it could be first or second.
The conduct of strategy, then, requires the competitive
arena to be "local," either in the literal, geographic sense
or in the sense of being limited to one product or a hand-
ful of related ones. The two most powerful competitive
96 HARVARD BUSINESS REVIEW
All Strategy Is Local
Efficiency in Place of Strategy
— — — — — — — — — — — — - ^ — _ . — ^ w « « « « « «
- — ^ _ _ _ ^ _ _ _ _ _ .
Companies can vary enormously in their operating efficien-
cies, and these differences can be sustained for many years.
But operating efficiencies are not a competitive advantage

37. because they can be, and usually are, adopted by other com-
panies. Also, competitive advantages are related to character-
istics ofthe external environment in which a firm operates-
primarily, its competitors-and not to its internal practices.
Take bar code scanning in the retail industry. The first
firms to install scanning equipment had a big advantage
overtheir slower competitors.They knew on a daily and ulti-
mately instantaneous basis what they had sold and thereby
gained better control of inventory and ordering processes.
But since the bar code systems were not proprietary to the
retailers (they had been developed and manufactured by
third-party firms that were all too willing to see them in-
stalled everywhere), the first movers did not sustain any ad-
vantage. A company's best and most innovative usesof infor-
mation technology, business models, financial engineering,
and almost everything else that applies to operations suffer
from the same availability to rivals. What a firm can do, its
competitors can eventually do as well. IT effectiveness, HR
policies, financiat strategies, and soon are essentially aspects
of what it means to operate efficiently. And operating with
superior efficiency is the only method of competing available
to a company that is separated from the conditions in which
strategy can make a difference.
nologies are highly stable, they eventually become avail-
able to all firms. Advantages based on customer captivity
are similarly perishable. Aside from literally passing away,
currently captive customers may move or age into new
markets.
Economies of scale can make up for these sorts of
losses. Coca-Cola's infrastructures, for example, enable the
company to attract more new customers, and to do so
more profitably, than its smaller and less-established com-
petitors can. Its weapons include more extensive adver-

38. tising and, thanks to scale advantages in distribution,
lower prices. Because of similar scale advantages, Intel
can spend many times as much as Advanced Micro De-
vices, IBM, or Freescale (a spin-off of Motorola) on de-
veloping new microprocessors and thus achieve domi-
nance with each new generation of its signal product.
Even when a rival has temporarily moved ahead, Intel (so
far, at least) has always had the time and the resources to
recover.
However, economies of scale must be accompanied by
some degree of customer captivity if they are to confer
sustainable competitive advantages. And with-
out such advantages, firms that have a domi-
nant share of their market wiil be forced to sur-
render some of it to new entrants. Even trivial
switching costs can enhance captivity and thus
multiply the advantages of scale. For example,
before the advent ofthe remote control, sheer
inertia kept fans ofa popular TV program from
abandoning whatever show came next, which
might have been one the network was trying to
launch. Now, the most sedentary couch pota-
toes will not hesitate to seek something more
to their liking. To their delight, their fondness
for choice has brought forth a proliferation of
program options; to the major networks' detri-
ment, it has spawned a greater number of com-
petitors and, hence, smaller viewerships.
Sustainable dominance is more likely in mar-
kets of restricted size. It is paradoxical but true
that economies of scale are subject to scale lim-
itations themselves. First of all, economies of
scale require levels of production above a cer-

39. tain size. Such scale is easier to attain in large
markets. Past a certain point, however, econo-
mies of scale cease being commensurate with
continued increases in quantity. In fact, they be-
come subject to diminishing returns, disadvan-
taging a larger competitor. In a restricted mar-
ket, by contrast, economies of scale are much
more difficult for a new entrant to achieve be-
cause it may have to capture 20% to 25% of the
market - a difficult threshold to reach when
each incremental gain comes out of the in-
cumbent's existing share. But unless the new
entrant reaches those levels, its economies will not come
close to paralleling the incumbent's.
The second reason that sustainable dominance is more
likely in markets of restricted size is that many fixed costs
are fixed only within the region or product market in
question. Expanding into another region that cannot be
served by an existing distribution infrastructure, for in-
stance, will necessitate new investment. To take another
example, economies of scale in advertising may be limited
to the area in which the language ofthe ad is spoken.
When a market gets too big, diseconomies of coordina-
tion can prevail over economies of scale. In expanding
markets, globalization has undermined profitability by
undercutting existing economies-of-scale advantages.
The story is told most clearly in manufacturing. When the
automobile industry was fragmented into national seg-
ments, each had room for only a small numher of highly
profitable participants-such as GM, Ford, and Chrysler, in
the United States, and Renault, Citroen, and Peugeot,
in France. With glohalization, these segments increas-

40. ingly coalesced into a single international market capable
SEPTEMBER 2005 97
All Strategy Is Local
of supporting a large number of competitors. A viable
share of this global market-that is, one offering absolute
scale advantages-was much easier to attain than a viable
share of a local market, which would have required gain-
ing a substantial market share. As a consequence, entry
and competition accelerated, to the marked detriment of
automobile manufacturers'competitive positions in their
home markets.
Scale advantages that endure in the face of increased
globalization are in markets limited enough to be domi-
nated by one or a small number of competitors. These
are the "local" markets, in geography or product space,
that the Microsofts, Intels, Ciscos, Coca-Colas, and Best
Buys have focused on, either by instinct or by design.
Wal-Mart and the
Retail Industry
Wal-Mart offers the most powerful demonstration ofthe
importance of dominating a local market. The retailer
began in the south-central region ofthe United States, ex-
panding steadily at the periphery of its territory. But it did
Supermarket Profitability and Local Market Share
Despite an increasingly global retail market, thinking locally
paid off for certain grocery

41. chains in the fiscal year ending in 2002. The most profitable
were the ones that dominated
their local markets. (Here, profitability is defined as pretax
return on invested capital.)
14% -
12% -
10% -
8% -
6% -
4%
2% -
0% -
Stater Bros.
Great A&P
•
1
Delhaize
•
• Safeway
• Kroger

42. •
Albertson's
• Winn-Dixie
„ • , . H.E. Butt
Publix
10% 20% 30% 40%
Local M a r k e t Share
Source: Accenture report Grocery Study: High Performance
Characteristia, September 2003
not stop there. It is now the largest retailer in the coun-
try-indeed, in the world.
Although we attribute Wal-Mart's historical perfor-
mance primarily to a strategy of local dominance, there
are competing explanations for the retailer's success.
Some observers have argued that Wal-Mart owes its su-
perior returns to its enormous size and, as a consequence,
its purchasing power. Alternatively, Wal-Mart is held up
as a model of operating efficiency, which, critics charge,
sometimes comes at the expense of its labor force.
But enormous size alone does not deliver a competitive
advantage. If the purchasing power that comes with size
were responsible for the company's success, then Wal-
Mart's profitability should have increased as the company
grew. Yet its operating margins (earnings before interest
and taxes) have not increased since hitting their high wa-
termark in the mid-1980s. In the years around 1985, Wal-
Mart had operating margins of 7% to 8% of sales. Recent
margins in its U.S. discount stores division have been about

43. the same. But with Sam's Club (Wal-Mart's warehouse
centers) and foreign operations included, overall margins
drop below 5%. Also, in the early 1980s, Wal-Mart was no
more than one-third the size
of Kmart and should have
suffered from a purchasing-
power disadvantage. Yet Wal-
Mart's margins at the time
were substantially higher than
Kmart's were. As Wal-Mart has
grown, however, its profit mar-
gins have suffered in com-
parison with those of more
geographically concentrated
competitors, such as Target.
The purchasing-power ex-
planation also defies economic
logic. At least 90% of Wal-
Mart's sales are made up of
nationally branded products
that are sold through a wide
range of competing outlets.
The producers of these brands,
by their own testimony, are
reluctant to favor one retailer
over others and risk antagoniz-
ing a majority of their distrib-
utors. As a result, they offer dis-
counts to Wal-Mart only to the
extent that Wal-Mart's more
efficient distribution systems
lower their own costs. Looked
at closely, purchasing power
does not seem to be chiefly re-

44. sponsible for the Wal-Mart suc-
cess story.
50% 60%
98 HARVARD BUSINESS REVIEW
All Strategy Is Local
Are superior operating efficiencies, then, the key fac-
tor? Certainly, Wal-Mart enjoys some advantages of effi-
c i e n c y - f o r instance, lower labor costs than those of
Kmart. But as with purchasing power, economics and the
broad historical record suggest otherwise. Greater oper-
ating efficiency should lead to greater profitability. If Wal-
Mart has a special talent for efficient operation, then that
strength should be apparent in all the company's divi-
sions. Yet Sam's Club appears to be no more profitable
than the other two major warehouse chains, Costco and
BJ's Wholesale Club. The fact that Sam's Club is the least
geographically concentrated ofthe three competitors ap-
pears to have offset any advantages derived from Wal-
Mart's efficiency. Even though competitors over the years
have copied many of Wal-Mart's cutting-edge techniques,
such as outsourcing to China and requiring leading sup-
pliers to put RFID tags on their goods, the deterioration
in the company's margins can be blamed on its inability
to replicate the same local economies-of-scale advantages
in the new regions it has entered. (The 2002 McKinsey
study "Retail: The Wal-Mart Effect" illustrates this point
in greater detail.)
Wal-Mart's experience overseas tends to confirm the
limited impact ofthe retailer's operating advantages. Be-

45. cause the operations and technologies of Wal-Mart's for-
eign competitors are less advanced than those of com-
petitors in the United States, the company should be able
to parlay this competitive edge into operating margins
abroad at least as high as those of its domestic operations.
In fact, overseas returns for Wai-Mart, whether on sales
or on invested capital, are less than half its domestic
margins. Especially in countries
like Germany, where Wal-Mart
faces entrenched competitors with
dominant local-market shares,
Wal-Mart's earnings performance
has been markedly substandard.
Our point is that while Wal-Mart's
operations may be more efficient
than those of its competitors, that
advantage loses its power in a for-
eign market dominated by a do-
mestic company.
Substantial, regionally deter-
mined fixed costs for advertising,
distribution, and store supervi-
sion provide the locally domi-
nant competitors with operating
cost advantages that most likely
overwhelm any differences in effi-
ciency that companies like Wal-
Mart obtain by applying widely
available retailing technologies.
In its discount store operations
within the United States, where
Wal-Mart is the one that benefits from local economies of
scale, the company is an almost irresistibly powerful com-

46. petitor. Overseas and even in the U.S. warehouse store cat-
egory, where others enjoy these advantages, Wal-Mart is
merely ordinary. Sam Walton's genius was to recognize
these facts first by establishing dominance in a core region
and then by attacking weaker competitors at the margins
of that territory, where his core advantages could be ex-
tended with relative ease.
What is true for Wal-Mart appears to be equally true
for other areas of retailing, including banking. In Jim
Collins's list of "good to great" retail companies, Kroger,
Wells Fargo, and Walgreens all had strong positions in
local or regional markets. The one retail company that
made Collins's list without being in such a position-Cir-
cuit City-has fallen on very hard times indeed. Moreover,
a systematic analysis of particular sectors shows a close
connection between local or regional market share and
profitability (see the exhibit "Supermarket Profitability
and Local Market Share"). And retailer-manufacturers
like Benetton that were the evangelists of a new wave of
global retailing have since largely retreated to their core
markets.
Pharmaceuticals and R&D
Pharmaceutical firms have been dramatic producers of
shareholder value throughout the past 20 years of glob-
alization. As this record unfolded, the industry's structure
changed to refiect the logic of specializing in particular
areas of research and the drugs that emerge from them
StPTEMBER 2005 99
All Strategy Is Local

47. JO CONFER SUSTAINABLE
COMPETITIVE_ADVANJAG_E_Ŝ
economies of scale must be accompanied by some
degree of customer captivity.
and to encompass a global network of local distribution
systems.
What has happened is that basic research has migrated
out of large pharmaceutical companies and into smaller,
more narrowly focused firms that specialize in research.
Roughly half of the licensed new drugs that big firms seek
to bring to market are licensed from these smaller research
companies, and this portion seems to be increasing.
With the expansion of global markets, such companies
can achieve scale advantages formerly the exclusive prop-
erty of large companies, given the size and expense ofthe
infrastructure required for major research. The result is
that large companies themselves-having lost their scale ad-
vantages - must now focus on particular product areas.
Another new development for big drug companies is
cross-border mergers-as we saw with Britain's Beecham
and the U.S. company SmithKline (before the merger
with the UK's Glaxo Wellcome), for instance, and with
Sweden's Pharmacia and the U.S. firm Upjohn (before
their acquisition by Pfizer). Cross-border mergers and con-
centration on particular diseases (such as Amgen's focus
on arthritis-not the company's only specialization) both
represent responses to the increasingly local imperatives
of global competition.
Globalization has eroded competitive advantages
among the established drug companies just as it did in
the automobile industry. Fortunately, the benefits of spe-

48. cialization by research area have allowed small drug firms
to seek, though not always find, competitive advantages
and operational efficiency within particular product mar-
ket niches. By acquiring licenses from these focused com-
panies, the major pharma firms are simply adapting to
tbe new strategic mandates that the advent of global mar-
kets has brought about.
In contrast to the development of new drugs, their mar-
keting remains an essentially local operation. Selling new
drugs through U.S. doctors, hospitals, and pharmacies
has always involved U.S.-based clinical trials, sales teams,
and distribution systems. Marketing is also targeted to
medical specialties. For a U.S. firm to carry out all these
functions in Germany, for instance, it would have to have
an elaborate infrastructure there; similar infrastructures
would be needed in all the other significant national mar-
kets. Each of these organizations would have a large fixed-
cost component as well. The patients reached by such
marketing efforts happen to be consistent in their pur-
chases, which translates into substantial customer captiv-
ity. As a result, each national drug-marketing organization
enjoys competitive advantages in both its geographic and
its specialty markets.
The efficient marketing of drugs, therefore, requires a
full range of national marketing organizations. Compre-
hensive global networks of locally dominant entities can
be formed by several means, including licensing, joint
ventures, and cross-border mergers. The recent wave of
transnational mergers is easily explained by the pres-
ence of competitive advantages based on local econo-
mies of scale.
Thus, the structure ofthe modern large pharmaceuti-

49. cal organization looks like a giant tree trunk connecting
sets of roots and branches. The drug research and devel-
opment, or "root," end is increasingly handled by firms
specializing in particular sciences and products, and the
distribution end is handled by strong local organizations,
either of the now merged pharma company or of its affil-
iates. Perhaps this trunk, through which specialized pro-
ducers pass their creations to equally specialized distrib-
utors, should replace "drug pipeline" as the industry's
defining metaphor.
Consumer Nondurables:
Coke and Pepsi
Producers of consumer nondurables constitute another
group of companies whose prosperity has withstood the
challenges of globalization. Companies such as Coca-Cola,
Colgate-Palmolive, Nestle, PepsiCo, and Procter & Gam-
ble, all of which were market value leaders 20 years ago,
have continued to produce high returns. The products
they sell have well-established global identities. However,
their relative competitive positions vary dramatically
across national markets. Local economies of scale in ad-
vertising and distribution are an important competitive
advantage for all these companies, especially when com-
bined with babit-based customer captivity. The geo-
graphic advantages these multinational corporations pos-
sess have allowed them to do a good job of defending
themselves against one another (although no domestic
company has stepped forward to challenge them).
Local strategic factors have always been an essential as-
pect of competition among these well-established com-
panies. But when Pepsi announced that it would chal-
100 HARVARD BUSINESS REVIEW

50. All Strategy Is Local
lenge Coca-Cola's global dominance, with the goal of
more than doubling its sales outside the United States,
it made the mistake of ignoring the local nature of the
markets in which it presumed to compete. Coca-Cola re-
sponded with a focused attack in the one market-Vene-
zuela-where Pepsi was the leader. Pepsi's position there
depended on its local bottler and distributor, which en-
abled Pepsi to realize economies of scale in advertising,
sales, support, and distribution. In 1996 Coca-Cola made
the bottling and distribution company an offer it could not
refuse, displacing Pepsi as its cola source and wiping out
Pepsi's strongest presence outside the United States.
Coke and Pepsi may be quintessential global brands,
but their competitive advantages, as Pepsi found out the
hard way, must be defended one local market at a time.
Telecommunications
and Media
In no other industry has the chasm between broad global
ambition and local success been as great as in telecom-
munications and media. The Internet, with its global
reach and ubiquitous presence, has been the protagonist
in the narrative of increasing global interconnectedness.
Satellites and other new distribution technologies, cou-
pled with the digitization of virtually everything, have
been widely expected to usher in a new era of universal
integrated content. Yet the companies in this industry
that have achieved high returns on capital and created
value for their shareholders have traditionally been-and
still are-those dominating local markets. Nothing seems
to have changed in this ostensibly new era.

51. In telecommunications, would-be global heavyweights
WoridCom and Global Crossing had bouts with Chap-
ter 11 bankruptcy protection. Traditional long-distance
competitors like Sprint and Qwest have had negative re-
turns on invested capital, little if any revenue growth, and
awful stock performance. Some have been absorbed by
local telephone companies, and others, namely Qwest,
have survived only by buying a regional Bell. Even AT&T,
once the dominant long-distance and international
communications firm, saw its performance deteriorate
steadily before being acquired this year by SBC (formerly
Southwestern Bell, one of the regional companies cre-
ated in the breakup of AT&T in 1984). In the United
States, the telecommunications companies at the head
of the pack after two decades of upheaval are former
local Bell operating companies-Verizon, SBC, Qwest, and
BellSouth.
The situation in Europe and Asia is similar to that
in the United States. The leading (as measured by prof-
itability and market value) telecommunications firms pro-
viding landline services, such as NTT in Japan, France
Tdldcom, Deutsche Telekom, and Telefdnica in Spain, all
have strong local franchises.
The same pattern holds for wireless communications.
The profitable operators in the United States are Verizon
and Cingular. Verizon's strength is in the Northeast; its
base consists largely ofthe former wireless subsidiaries
of NYNEX and Bell Atlantic. Before Cingular acquired
AT&T Wireless, Cingular's customers came mostly from
the wireless operations of BellSouth and SBC-again, re-
gionally based organizations. The more nationally ori-
ented providers, AT&T (whose acquisition by SBC awaits
regulatory approval) and Sprint, have fared poorly. The

52. only successful national competitor has been Nextel,
which has specialized in business communications and
offers a walkie-talkie service with its phones. In Europe,
the only company with strong positions in more than its
host country is Vodafone, which has a major share in the
United Kingdom and some other markets. Otherwise,
the field is populated by national champions.
In media, broadly defined, actual experience has been
even more strikingly at odds with prevailing strategic wis-
dom, which in the last ten to 20 years has proclaimed
that successful media companies would be those that in-
tegrate content and distribution, are global in reach, and
embrace and master new technologies. The premier com-
panies pursuing these strategies have been four U.S.-
based media giants: Time Warner, Viacom, Disney, and
News Corporation (which was originally based in Aus-
tralia). One European company that followed this path,
Vivendi Universal, imploded spectaculariy, and another,
Bertelsmann, has pulled back from America. But the
American companies have also stumbled. In the past ten
years, they have all been able to grow revenue, but their
top-line growth has not translated into substantial value
creation. None of the leading global media companies
has equaled the performance of the S&P 500 over the
past 14 years; their average has been lower by almost 5%
per year.
This performance history is in sharp contrast to that of
the old-fashioned, locally based newspaper companies in
the United States. These companies have not grown their
revenues as fast as the big media firms, which is under-
standable, given the dated nature of their products. How-
ever, their shareholders' returns have generally exceeded
those of the broad market indexes. Their strategies, fo-
cused on dominating their local markets, have yielded far

53. greater returns than those of the big media companies.
(See the exhibit "More Isn't Always More.")
The economics underlying these experiences in both
the telecommunications and the media industries should
by now be familiar. Landline telecommunications, cel-
lular phone systems, and local newspapers all involve
significant fixed costs within each regional market, which
are a requirement for economies of scale. These econo-
mies have created barriers to entry, protecting the in-
cumbents. Potential entrants would have to seize suffi-
cient local market share to become viable competitors.
SEPTEMBER 2005 101
All Strategy Is Local
and the incumbents' existing degree of customer captivity
has made this difficult to achieve. By contrast, global mar-
kets for long-distance telecommunications, film produc-
tion, recorded music, and books are so large that they will
support many entrants, each with a relatively limited
market share. As a result, these industries lack effective
barriers to entry, must cope with intense and uncontrol-
lable competition, and suffer from disappointing prof-
itability and shareholder returns.
Information Technology
The history of distributed personal computing illustrates
the importance of concentrating on narrowly defined
product markets in establishing competitive advantages.
In the early 1980s, at the dawn of the PC era, a number of
large, well-financed companies were in command of the
technologies that are now at the core of modem infor-

54. mation processing. Apple and IBM, early leaders in the
market, demonstrated their abilities as developers of
software, hardware, and microchips. Digital Equipment
was a leader in time-share computing, the precursor to
modem distributed-computing networks, and in Ether-
net connectivity technology. Xerox, with its Palo Alto Re-
search Center, was a pioneer in software technology, and
the company enjoyed a strong marketing presence at the
office level, where much PC equipment was purchased.
AT&T was a leader in digital communications, systems
software (the UNIX system was AT&T's creation), semi-
conductor technology, and fiber optics. Motorola had
well-developed capabilities in chips and communications.
Hewlett-Packard was strong in a wide area of individual
computing technologies and incubated many of the lead-
ing technologists in Silicon Valley. Yet, with the excep-
tions of HP in the specialized market of printers and IBM
in enterprise applications software, none of these giant
companies is a significant player in today's information
technology world.
Instead, competitive advantages and the value creation
they spawned have been in the hands of companies that
took a far more local approach to product development.
Microsoft began by focusing narrowly and obsessively
More Isn't Always More
A common strategy among U.S. media giants has been
to expand, both in geographic reach and in products of-
fered. The big four have delivered revenue growth over
the past decade-but, as their low shareholder returns

55. show, they haven't managed to generate value.
They could take a lesson from U.S. newspaper com-
panies, whose shareholder returns have, in general,
beaten market indexes. The key to the newsies'success?
Domination of local markets.
BIG MEDIA COMPANIES
Company
Time Warner
Viacom
Disney
News Corporation
Annual
Revenue
Growth
1994-2004
21.3%
16.0%
13.4%
11.4%

56. Annual
Shareholder
Returns
1991-2004
1.4%
5.8%
8.3%
7,8%
Average
S&P 500
15.5% 5.8%
10.5%
TRADITIONAL NEWSPAPER COMPANIES
Company
Tribune
McClatchy
Washington Post
Gannett
Scripps
New York Times
Knight Ridder

57. Pulitzer
Average
S&PSOO
Annual
Revenue
Growth
1994-2004
11.6%
10.5%
8.3%
7.7%
6.4%
3.8%
1.5%
-1.0%
6.1%
Annual
Shareholder
Returns
1991-2004
13.3%

58. 13.9%
13.7%
14.1%
12.6%
11.5%
9.7%
13.8%
12.8%
1O.S%
Source: Value Line. Both tables list shareholder returns for
1991 through 2004 but include revenue growth only for the past
ten years. The reason for
using a later starting point to track revenues is to ensure that
revenue growth rates were not built into the share prices at the
start; high rates of reve-
nue growth might have lowered subsequent investment returns
by raising the initial share prices.
102 HARVARD BUSINESS REVIEW
All Strategy Is Local
NONE OF THE LEADING GLOBAL MEDIA COMPANIES
has equaled the performance of the S&P 500 over the

59. past 14 years.
on the PC operating system, designing its early word-
processing, spreadsheet, and browser software to protect
and extend that franchise. Intel concentrated solely on
chips and, after the mid-1980s, microprocessors. Cisco spe-
cialized in routers and other intracompany network sys-
tems, incorporating both hardware and software. Dell ini-
tially devoted itself entirely to personal computers sold
directly to customers, bypassing established and, it proved,
less efficient channels. Even IBM and HP have been suc-
cessful in "local" rather than general markets. Firms with
strategies like Apple's, designed to dominate the PC mar-
ket as a whole, have not succeeded, in the new industry of
personal-computing networks, successful companies have
confined themselves to local product markets.
'Rvo factors account for this outcome. First, economies
of scale apply within particular segments, not to the in-
formation technology market as a whole. Network ef-
fects, through which customers receive greater value as
more users acquire the same products or technology,
are specific to individual segments. Those accruing to
users of operating systems, for example, don't spill over
to users of communications software. These effects have
contributed significantly to the leading positions of
Microsoft and Cisco in their respective markets. Large
fixed development costs are characteristic of both soft-
ware code and microprocessor design and production. By
adding features and capabilities to successive generations
of their basic products, Microsoft, Intel, and Cisco have
managed to distribute those costs across a greater num-
ber of unit sales. Since all three companies enjoy power-
ful customer captivity and a dominant market share, they
can in turn afford to spend much more on the fixed costs
necessary to produce the next generation of technology,

60. yet they will still have lower costs per customer than their
rivals, an advantage that helps them maintain their dom-
inance. Apple's recent decision to switch to Intel micro-
processors underscores the power ofthis advantage.
For a company like Dell in PC manufacturing, a com-
modity business that is not evolving much, development
costs are far less significant, meaning that economies of
scale are also less important. Customer captivity is also
considerably weaker in the interchangeable world of PC
hardware. Although Dell has tried to induce habit forma-
tion and boost switching costs among its institutional cus-
tomers through ordering systems that are tightly in-
tegrated with production, evidence suggests that its
customers are far less attached to its products than Micro-
soft's, Intel's, and Cisco's users are to theirs.
For Dell, the primary benefit of its narrow product
focus-until recently, only PCs-appears to have been sim-
plicity and clarity, which have allowed Dell to concentrate
on operational efficiency. Compaq, the most challenging
competitor in Dell's early years, seemed to have similar
success after it refocused itself in 1991 to produce generic
PCs as efficiently as possible. But Compaq lost this clarity
of vision. It acquired first Tandem and then Digital, and its
performance deteriorated. Clarity and simplicity-espe-
cially in markets without barriers to entry, where opera-
tional efficiency is everything - are two of the greatest
benefits that a local focus imparts.
Keeping It "Local"
For all the talk ofthe convergence of global consumer de-
mand, separate local environments are still characterized,
in both obvious and subtle ways, by different tastes, dif-
ferent government rules, different business practices, and

61. different cultural norms. (The single most glaring excep-
tion may be in luxury goods, where brands like Prada and
Louis Vuitton have outlets throughout the developed
world. These products have global appeal for the special
category of cosmopolitan, high-income consumers.) And
as our comparison of vertically integrated media and
newspaper companies makes clear, the decision to con-
centrate in a narrow set of products or services has its own
benefits. Coping with either regional differences or an un-
wieldy range of offerings puts heavy demands on any
company's management.
The more local a company's strategies are, the better
the execution tends to be. Localism facilitates decentral-
ization-and since the days of Alfred Sloan, decentralized
management has consistently served as a superior struc-
ture for concentrating management attention. Decen-
tralization matters for both product space and physical
territory. GE has always been noted for its stock of man-
agement talent, but the efficiency with which it deploys
that talent is equally important. This efficiency can be at-
tributed to a decentralized organizational structure: The
company's many activities are organized into indepen-
dently focused divisions with clearly formulated, local
strategic objectives, such as the need to be first or second
in the relevant industry segment.
SEPTEMBER 2005 103
All Strategy Is Local
Another powerful illustration ofthe virtues of concen-
tration is the performance of Microsoft, whose remarkable
success is built primarily upon two related types of soft-

62. ware, versus that of Apple, which has never stopped striv-
ing to excel in software, hardware, and media products
but has enjoyed only intermittent successes mixed with
frequent disappointments. Apple's current profitability is
attributable to the iPod, not the personal computer.
Strategies that are local in the nongeographic sense im-
prove companies' competitive strength by facilitating co-
operation across product boundaries. If, like Apple, Intel
had decided to produce computers and software as well as
CPUs, it would clearly have had much more difficulty
forging its partnership with Microsoft, a relationship that
has contributed so heavily to Intel's dominance of its own
industry. Intel's skill at designing and producing micro-
processors and Microsoft's at writing software constitute
a joint enterprise of exponential efficacy.
With the globalization of manufacturing has come an
increase in competition, along with a decline in prof-
itability. Companies and countries that ignore this reality
and try to compete in global markets for manufacturing
face stagnation and poor performance, not to mention
the challenge of going up against billions of capable, low-
wage Chinese and Indian workers. The countries that
have tried to follow this p a t h - m o s t notably Japan, Ger-
many, and France - are suffering the consequences of low
economic growth and underemployment.
At the same time that manufactured goods (even as
they increase in variety, quality, and functionality) repre-
sent a shrinking portion of people's consumption bud-
gets, especially in the developed world, services of all
kinds, including necessities like medical care and desir-
ables like entertainment, represent a growing one. Be-
cause services are more often than not provided locally,

63. their ever-increasing fraction of countries' gross domes-
tic products could create the conditions for a renais-
sance in another local pursuit; the making of corporate
strategy. ^
Reprint R0509E
To order, see page i6i.
'The only difference between you and me, Flanders,
is that I read the homework before eating it."
104 HARVARD BUSINESS REVIEW
Harvard Business Review Notice of Use Restrictions, May 2009
Harvard Business Review and Harvard Business Publishing
Newsletter content on EBSCOhost is licensed for
the private individual use of authorized EBSCOhost users. It is
not intended for use as assigned course material
in academic institutions nor as corporate learning or training
materials in businesses. Academic licensees may
not use this content in electronic reserves, electronic course
packs, persistent linking from syllabi or by any
other means of incorporating the content into course resources.
Business licensees may not host this content on
learning management systems or use persistent linking or other
means to incorporate the content into learning

64. management systems. Harvard Business Publishing will be
pleased to grant permission to make this content
available through such means. For rates and permission, contact
[email protected]