ness decision, or that the selected providers, or the selected country, are not appropriate.
Once the provider is selected, the contract terms need to be ļ¬nalized. Legal and con-
tract issues are described in detail in Chapter 6, with a few key items noted here:
ā— Pricing and additional costs. The key issue here is to have a full understanding of
which costs will be covered by the provider for items, such as travel to the client
site, software licenses, training, and hardware acquisitions.
ā— Issues of intellectual property. warranties, and conļ¬dentiality.
ā— Incentives and penalties. These are the mechanisms by which the providerā™s
performance is better aligned with the customerā™s goals.
In general, the contract is secondary to building a relationship. Long and overly detailed
contracts are often an indication that there is little trust and that relationships were not
built during communications with the provider. Such cases do not foretell a successful
offshore project. Understand that the provider also has constraints and needs, such as
68 The fundamentals
the set-up time for the infrastructure and the project team, investments in training, and
investment recovery. An environment must be created where the business interests of
the two parties are in alignment over the life of the relationship. Relationships may also
be important with those ļ¬rms that were not selected; contact them and explain the rea-
sons why they were not selected, since you may want to work with them in the future.
You are now ready to embark on the offshore journey. Bon Voyage!
The principal mistake, as companies begin their offshore journey, is to underestimate the
importance of careful preparation. Sufļ¬cient time and budget must be available for laying
the foundation, provider identiļ¬cation, and the ļ¬nal selection.
Create a small, agile but powerful project launch team with members who are open to
offshoring. If you cannot do this with internal resources, consider buying knowledge from
Internal resistance is the greatest barrier to offshoring. Visions and beneļ¬ts must be
communicated clearly early in the journey. Consider change management approaches to
address internal resistance.
A low-risk pilot project is recommended. It can be used to test if your company is ready for
offshoring, to test a certain country, or to assess a new provider.
Be aware that the sequence of country-ļ¬rst or provider-ļ¬rst is dependent on several key
factors. If the engagement is short, then choosing by provider is probably best. For a longer
engagement, country selection should come ļ¬rst. The country choice can be the result of
investigation, but it is legitimate to make the choice because of personal contacts or
personal reasons. Select a location your key people will enjoy traveling to.
The criteria for provider selection criteria are different for every client organization. Be
selective in sending out RFIs and RFPs to potential providers; long lists are a waste of time.
Do not rely solely on the RFP responses. Due diligence, the careful investigation of the
provider, is critical. Without such validation, the providerā™s weaknesses will not be
discovered until a later stage. A site visit abroad is recommended in case of large or
complex projects or if long-term collaboration is desired.
The contract is secondary to building a partnering relationship. Try to create an environment
where the business interests of the two parties are in alignment over the life of the
4 The offshore country menu
Close to 100 nations are now exporting software services and products. The āoffshore
menuā is immense, with many nations to satisfy any taste. These nations span the eco-
nomic spectrum from newly industrialized economies, through transition economies,
to developing economies, and even some least-developed nations.
The 1990s was the ļ¬rst decade of softwareā™s true globalization.1 The 1990s saw the
rise of three celebrated success cases, the āthree ā˜Iā™sā ā“ India, Ireland, and Israel. These
were the three nations that seemed to appear overnight as global centers of important
software activities. One of the most interesting features of the āthree ā˜Iā™sā is that each
of these three nations developed and specialized in different aspects of software. That
is, each one progressed to become a global software player in different ways: India in
offshore programming, Israel as an incubator of software products, and Ireland in pro-
gramming services and localization services.
Three tiers of software exporting countries
The G7 nations2 produced much of the worldā™s software in the ļ¬rst few decades of the
computer era. High-tech exporting used to ābelongā to these nations with the USA as
the hegemonic power in software. Until roughly 1990, very few nations exported soft-
ware products or software services at any non-trivial levels, including that which today
we call āoutsourcingā. The G7 nations are still at the core of the Tier-1 software nations
(see Table 4.1), the Mature Software Exporting Nations. These G7 nations have a tra-
dition of exporting high-technology and knowledge-intensive products and services. In
particular, the USA (with its giants, Microsoft and IBM) continues to dominate world
markets. The other G7 nations, Japan, Great Britain, Germany, France, and Canada,
have had successful software (and computer hardware) industries spanning many
decades. The one outlier among the G7 is Italy, which has never developed a strong
software sector for an economy of its size. To the Tier-1 software nations we add
several other advanced industrialized nations: The Netherlands, Sweden, and Finland,
which have all had strong software export sectors.3
To Tier-1 we also need to add the āthree ā˜Iā™sā ā“ India, Ireland, and Israel. All three
nations have developed robust software export industries (we return to the three āIās
70 The fundamentals
Table 4.1 The 3-tier taxonomy of the worldā™s roughly 100 software exporting nations4
Tier-1 Mature software Mostly industrialized nations such as: USA, Canada,
exporting nations UK, Germany, France, Belgium, The Netherlands,
Sweden, Finland, Japan, and Switzerland
Entrants from the 1990s: Ireland, Israel, and India.
Entrants from the 2000s: China and Russia
Tier-2 Emerging software Brazil, Costa Rica, Mexico, The Philippines, Malaysia,
exporting nations Sri Lanka, South Korea, Pakistan, Ukraine, many
other Eastern European countries, and several
Tier-3 Infant stage software Cuba, El Salvador, Jordan, Egypt, Bangladesh,
exporting nations Indonesia, Vietnam, and 10ā“20 others
Non-competing Non-competing About 100 of the mostly, small, least-developed
countries of the world, including most African, and
many Middle-Eastern nations. These nations have
few to no software exporting ļ¬rms
later in this chapter). Finally, we add the two newest entrants to the Tier-1 nations:
China and Russia. As recently as 1999 it would have seemed far-fetched to classify
Chinaā™s software export industry in Tier-1. But Chinaā™s software industry has been
maturing so quickly in the early 2000s that this is no longer debatable. Russiaā™s place
in Tier-1 may still be marginal at this point.
The software exporting nations are classiļ¬ed into tiers based on three criteria: industry
maturity, clustering, and export revenues. These criteria are soft criteria, and the tiers
themselves are constantly changing. They will surely be deļ¬ned differently a decade
ā— Industry maturity connotes the nationā™s tradition of exporting software. Most Tier-1
nations have been exporting software since well before 1990, with Russia and China
as the only exceptions. Tier-2 nations have been exporting since at least the mid 1990s.
ā— Clustering connotes some critical mass of software enterprises participating in the
software export industry. Tier-1 nations have hundreds, and in some cases
thousands, of ļ¬rms exporting software products and services. Clustering also
connotes a maturing collection (agglomeration) of secondary services to support
software companies, including consultancies.
ā— Export revenues are the magnitude of national software exports. Some
representative numbers: India 12.5 billion USD (2004), Israel 2.7 billion USD (2003),
Brazil 200 million USD (2001), Vietnam 30 million USD (2003), and Indonesia 30
million USD (2000).
Tier-2 nations are the Emerging Software Exporting Nations. All of these nations
already have signiļ¬cant software export industries, exporting at 25ā“200 million USD
per annum. Most of these nations have clusters of technology ļ¬rms either in major
71 The offshore country menu
metropolitan areas or in designated technology parks. These nations have dozens of
organizations exporting software, but usually less than 100. We use the neutral term
organizations because the unit that is exporting software may be a software subsidiary
of a multinational enterprise, or a home-grown, independent software company.
Most of the Tier-2 nations are unlikely to move up and join the mature Tier-1 software
nations. Their ļ¬rst liability is a small population base, which restricts their ability to
grow large industries. A second liability is unfavorable conditions, such as political
instability or immature stage of economic development. The strongest of these Tier-2
software nations, Brazil, Mexico, South Korea, and the Philippines, may coalesce and
form an intermediate second tier within a few years, and separate and distinguish
themselves from the smaller less robust countries in this tier. These more vibrant software
nations are the larger nations that possess the wealth and large labor pool of educated
human capital that is needed for growth.
Tier-3 nations are Infant Stage Software Exporting Nations with an insigniļ¬cant
impact on the global software market. Some Tier-3 nations have beneļ¬ted from some
foreign direct investment (FDI) in a number of their ļ¬rms, but it has been small, and
isolated to just one or a few ļ¬rms. Indiaā™s remarkable success is well-known, and gov-
ernments in a number of these nations have woken up to the potential economic beneļ¬ts
of software exports. They are working to encourage the software export sector. This is
discussed in greater detail in Chapter 10.
The software industries in Tier-3 nations are mostly ācottage industries,ā where compa-
nies are small and management is not professionalized. Transforming these industries will
take years. Most Tier-3 nations will not move up to Tier-2 because of their relatively small
size, which restricts their ability to grow large industries. More signiļ¬cantly, due to their
stage of economic development, or political instability, the industry growth will be stunted.
What country to choose?
In this section we look at the factors that determine where companies locate their soft-
ware work. After we cover a general list of criteria, two speciļ¬c locational factors are
discussed in greater detail: risk and government incentives. This is followed by a case
study of a small American ļ¬rm that was faced with a country choice decision and took
an interesting approach to solve it: trying several countries.
The many factors to consider in location decisions
What are the factors that should be in your companyā™s location decision? We begin
with four high-level factors.6
1 The type of activity that is going abroad. That is, whether the activity is basic
research, applied research, or development. For basic research the sites should be
72 The fundamentals
located next to major national engineering universities, as most talented people
are likely to be there. For basic research location costs and wages are less
2 Duration of engagement. Managers beginning the offshore journey sometimes
ask: āwhat comes ļ¬rst, choosing the country or choosing the provider?ā If the
engagement is short ā“ if it is one project (out-tasking) ā“ then choosing by provider
first is probably best. However, for a longer engagement, and certainly for any
kind of captive center (subsidiary), country choice should come ļ¬rst. The company
has to gain a deep understanding of the nationā™s context in order to choose the
3 The ļ¬rmā™s geographic orientation. That is, whether the ļ¬rm has regional or
global preferences, or some other afļ¬nity. The Swedes called geographic
orientation āpsychic distanceā to explain where their companies choose to locate
their research and development (R&D) centers abroad. Americans call this
ācultural distance.ā Geographic orientation explains why the British locate in
India, the Spaniards in Latin America, and so on. One research study even
measured the cultural distance between countries using a composite measure of
cultural characteristics7 in order to determine whether culture has an impact on
foreign investment ā“ it did.
4 Motivation. This refers to whether the ļ¬rmā™s motivations are supply- or demand-
driven. Supply-driven means that the ļ¬rm wants to have access to resources such
as high-end labor, low-cost labor, or technology know-how.8 Of course, access to
resources has been the dominant consideration for offshoring. Companies want
access to cheap and vast pools of labor. Demand-driven means the ļ¬rm wants
access to local (offshore) markets. For example, China has drawn nearly every
major high-technology ļ¬rm to its cities for this reason. Companies gain proximity
to important customers and markets ā“ and in the case of China, easier access
through governmental controlled markets. The other element of demand-driven
motivation leads companies to locate where they can better redesign products for
local use ā“ localization.
The demand for resources often leads companies to locate in geographic āclusters.ā
For example, of all foreign investment in India, Bangalore alone took more than one-
third (38% in 2002).9 At ļ¬rst, it seems counter-intuitive that foreign companies all
come in from afar to tap resources and cluster next to one another, sometimes setting
up shop in the same ofļ¬ce park.
Yet, there are many good reasons to cluster together. Accessing the labor pool is eas-
ier in clusters: the top engineers and scientists are more likely to be in the cluster or
move there. Companies want to be next to the best recruiting sources, and so they locate
next to scientiļ¬c competence centers, universities, or research institutes. Of course,
this also affords them proximity to leading research activities. Furthermore, companies
want to locate where, as one manager said to us: there needs to be a ābuzz.ā Being close
73 The offshore country menu
to major competitors is actually a good thing because it allows the ļ¬rm to gather soft
information about what competitors are doing. Finally, of necessity, in developing
nations with their poor infrastructures, one cannot ļ¬nd support services outside the
Missing from the above list of four factors is country specialization. Why? Shouldnā™t
technology executives make location decisions based on special national expertise in a
desired ļ¬eld? After all, one thinks of France for wine, Japan for cars, and Switzerland
for chocolate. In all these nations there are clusters of ļ¬rms that specialize in producing
or supporting each of their specializations. Are there no national specializations within
software? The short answer is that there are no nations in Tiers 2 or 3 that have coun-
try software specializations. All of these countries have smart people, but they are offer-
ing generic skills. These nations compete with one another via other national factors,
not software specializations. In some nations there may be one or two ļ¬rms that have
world-class capability in one dimension, but this does not constitute a national spe-
cialization. If a company is searching for very specialized expertise, it usually exists
offshore, but rarely as a national distinction.
Let us take a closer look at the supply-driven factors, as these are most important for
offshoring software in most cases. A.T. Kearney, an American consulting company,
compiled an offshore location attractiveness index summarized in Figure 4.1. The
index is made up of three criteria groupings to rate and rank nations qualitatively for
offshore IT work:
ā— Costs, which includes wages, infrastructure costs, tax, and regulation costs.
ā— Labor, which includes business experience, labor force availability, education and
language, and attrition.
ā— Business environment, which includes investor rankings, country infrastructure,
cultural adaptability, and intellectual property (IP) protection.
Notice that weights in Figure 4.1 were chosen by the studyā™s authors, while the weights
for your ļ¬rm are likely to be different. For example, cost may be weighted more heavi-
ly at your ļ¬rm. In fact, the country decision-maker should see in these data a classic
trade-off: those countries with fewer points on costs, such as Canada and Australia,
have many points on āBusiness environment.ā And, vice versa, those nations with
lower costs tend to have fewer points on āBusiness environment.ā
The āsurprisesā on this 2004 index, as the authors note, are the nations ranked imme-
diately after India and China, namely, Malaysia, the Czech Republic, and Singapore.
Malaysia, a country of 22 million people, is most noted for the massive government
project in the Multimedia Super Corridor stretching from the capital of Kuala Lumpur
to the new airport south of it. This geographic cluster includes the smart cities of
Cyberjaya and Putrajaya, with their excellent infrastructure in connectivity and facili-
ties. Malaysiaā™s potential may well lie in IT-enabled services (ITES) as much as soft-
ware. And due to its largely Moslem composition, the nation is becoming a preferred
destination for the many nations of the Islamic world.
74 The fundamentals
0 2 4 6 8
Costs Business environment Labor
Figure 4.1 Offshore Location Attractive Index (2004). Source: A.T. Kearneyā™s Location
One of the location criteria is country risk (introduced in Chapter 2). Offshoring is
about sourcing mostly from developing and emerging countries, which have historically
been more volatile, less stable, less predictable, and less transparent. When offshoring,
companies are exposed to increased risks of war, terrorism, rioting, uprising, conļ¬sca-
tion, expropriation, and currency crises. Thus, the consequences of country risk can be
severe because they affect, among other issues, business continuity, which is the ability
75 The offshore country menu
of the ļ¬rm to continue its core operations. Country risk is an important, sometimes
dominant, issue. Larger ļ¬rms diversify country risk by setting up operations and back-up
sites in a number of countries, rather than just one location. Thus, in event of a crisis,
they can shift operations.
There have been several tremors since the offshoring era began, such as the near
war between India and Pakistan in June 2002, the September 11th 2001 attack in the US,
and the period leading up to the American invasion of Iraq in early 2003. All of these
tremors heightened uncertainty, resulting in some delays in decision-making and some
project delays. But none of these tremors have seriously impacted offshoring. If a
limited Indianā“Pakistani war were to break out, it would disrupt a companyā™s business
by delaying some work and hindering travel, while a more severe war would have
long-term consequences on a companyā™s continuity if key systems and processes are
in India. Note that war is considered a āforce majeureā event, and in the event of
war, if the provider fails to meet commitments, it is not considered as breach of
The risk of conļ¬‚ict may well be overstated for many offshoring decisions.11 First,
critical production systems are usually not offshored. Second, software work tends to
be somewhat isolated from the rest of the economy and needs little in the way of
supply. Indiaā™s software industry has seen only insigniļ¬cant disruptions with all its
tensions with Pakistan. Similarly, Israelā™s large software technology sector has not
seen work disrupted by the nationā™s ongoing problems as a result of the Palestinian
Even Yugoslavia, during its 1999 war, as NATO jets bombed, saw software opera-
tions continue. When the war broke out, Clockwork, an Amsterdam-based Internet
services company, had just set up an ofļ¬ce in Novi Sad, Yugoslaviaā™s second largest
city. Several strategic targets in the city were destroyed during the attacks, such as its
TV station, the oil reļ¬nery, and the bridge crossing the Danube. The residential area
where the ofļ¬ce was located was not a target, and the team of ļ¬ve programmers was
able to continue the work. Broadband Internet connection continued to be available
and communication with Amsterdam was not disrupted.
Country risk has a number of secondary implications. One of these is the risk of gov-
ernment regulatory changes (also called sovereign risk). A generation ago it was com-
mon to fear conļ¬scation, nationalization, and expropriation. The Indian government,
in 1977, actually made business so uncomfortable for IBM at the time that IBM pulled
out of India entirely. It is hard to imagine these kind of events in todayā™s global mood,
but they may still happen. Less acutely, governments may change tax or subsidy struc-
tures that once favored offshore operations, making them less attractive. Thus, favor-
able tax treatments in, say, the Philippines or China may change. Governments may
also change the regulations that govern technology joint ventures in order to favor the
76 The fundamentals
Incentives provided by national governments
Countries compete with one another to attract foreign investors into their countries,
states, provinces, regions, and cities. The competition for such Foreign Direct Investment
(FDI) has become more intense in recent decades, with no signs of letting up. Com-
panies that want to build software centers are especially welcome (see Chapter 10).
Governments induce FDI in software activities through three principle incentive
ā— Governments reduce or eliminate various taxes. This can be done through a tax
holiday (such as a 5-year tax exemption), or by import-duty exemptions.
ā— Governments subsidize certain activities and investments. Some investors will
receive grants, although these are usually negotiated. A more common approach is