1. R&D INTERNATIONALIZATION AND CHOICE OF COUNTRY: CONCEPTUAL
MODEL AND APPLICATION
ANDRÉA GOMIDES
School of Economics, Business Administration and Accounting - University of São Paulo
Av. Rio Branco, 45 sala 1006 – Rio de Janeiro – RJ - CEP: 20.090-003
andrea_gomides@hotmail.com
EDUARDO VASCONCELLOS
School of Economics, Business Administration and Accounting - University of São Paulo
Av. Prof. Luciano Gualberto, 908-Cidade Universitária – São Paulo-SP – CEP: 05508-900
epgdvasc@gmail.com
ERWIN REZELMAN
SAP LABs – SAP AG
Av. Unisos, 950 – São Leopoldo – RS – Brazil - CEP: 93022-000
Erwin.rezelman@sap.com
Global competitiveness is influenced by the capacity to generate sustainable
innovation – that is, innovation that adds value for clients, creates a preference for the
company’s products and services, and brings results that permit compensation of all other
stakeholders. Finding the appropriate degree of R&D decentralization, taking into account the
characteristics of each business, is an important factor in obtaining these results. There is a
wealth of literature on the advantages and disadvantages of decentralization and the factors
that condition the decision to decentralize. When this decision is made, it entails a question of
key importance that has been much less widely studied: how does the company choose the
country that will host its new R&D unit? Seeking to answer this research question, we
developed a conceptual model based on a review of the literature, then selected a company on
which to test this model and conducted a case study. The chosen company was SAP, a global
organization that had to choose between four Latin American countries to host its new R&D
unit. We applied our model to their decision process, although the decision had already been
made at the time the present study was conducted. Our results demonstrated the utility of the
model for country choice decisions. Additional studies are required to validate the conceptual
framework and its applicability to other companies.
Keywords: R&D, Internationalization, choice of country, SAP AG, innovation
1. Introduction
Technological advance is no longer simply a means of standing out. It has become a
guarantee of activity in an increasingly competitive market. Some companies choose to keep
their R&D function centralized, generally at the organization’s headquarters, but when they
consider competition and the need for speeding up the process of developing products and
taking them to market, the centralization strategy is called into question, and arguments arise
for the implementation of a decentralized model. There is a wealth of literature on the
advantages and disadvantages of R&D internationalization, and on the trends followed by
this process. There are few studies on models used to select the location of a new unit. The
research question we sought to answer was which factors should be considered when
selecting the most appropriate country for a new R&D unit?
1
2. We found the answer at SAP, the world leader in enterprise resource planning software
and the first of the world’s major management companies to invest in Brazil. A model for the
choice of R&D center locations was outlined on the basis of a literature review and
interviews with SAP Labs’ (this is the name for SAP R&D centre) president for Latin
America. SAP’s decision to place its new R&D center in Brazil, rather than Argentina,
Mexico or Chile, then served to illustrate application of our conceptual framework.
2. Methods
We selected the case study as our research method. After an extensive literature
analysis, we idealized a conceptual model and devised an interview script. Two in-depth
interviews were conducted with, Erwin Rezelman, the President of SAP Labs Brazil. Due to
his interest in the subject and his countless contributions to our study, he was invited to co-
author the present article, reviewing the text and providing access to internal documents. The
use of diverse sources of information was instrumental in ensuring greater information
accuracy, in accordance with the triangulation principle (Eisenhardt, 1989; Yin, 1994).
The case study is an appropriate method when the context is complex and involves
many variables (Easton, 1995; Dubois & Gadde, 2002; Eisenhardt, 1989; McCutcheon &
Meredith, 1993; Miles & Huberman, 1994; Yin, 1994). SAP was selected because it went
through the process of selecting a country to host its new R&D unit.
Data was collected and analyzed over a six-month period. Both in-depth interviews
with Mr. Rezelman were conducted within this period, following a semi-structured interview
script devised after a conceptual model based on our literature search. When the final model
was ready, a table of factors, grouped into three macro factors (market in each country, R&D
inputs, and country profile) was devised and filled in by SAP.
3. Centralization and international decentralization of R&D activities
Centralization and decentralization of R&D structures has long been the subject of intense
discussion by several authors. The specificities of each company and business model mean
there is no single model of centralization or decentralization to be adopted. The factors we
will introduce should be analyzed by each company and checked against their specific
decentralization needs to help choose the best location for a new site.
For many years, R&D sites were only present in developed countries; they were
particularly widely concentrated in Europe, the U.S., and Japan. Companies in these countries
either kept their units in their home country or expanded into other major markets, which
required customization. Over the past ten years, this scenario has changed, as global
companies identified investment opportunities in countries that could potentially constitute
new markets, not only for the possibility of sales growth, but because the companies realized
that knowledge was no longer centralized at the great economic powers, but rather spread out
around the world.
In a 2006 survey, Booz Allen Hamilton and INSEAD asked 186 companies in 19
countries – with pooled R&D investments of over US$ 76 billion in 2004 – where they would
2
3. implement or expand innovation/R&D sites over the next 5 years and why. One finding of the
survey was growth in the percentage of companies that had decentralized their R&D sites, up
from 45% in 1975 to 66% in 2004.
Figure 1: Growth in foreign R&D sites
Source: Booz Allen Hamilton and INSEAD, 2006.
As for the countries that would receive investment over the next 5 years, the survey
showed that: 95 new sites were to be installed in China; 76 in India; 75 in the United States,
which remained as a key recipient of investment; and 11 in Brazil, which came next to last,
but was the only Latin American country to make it into the top 10.
Figure 2: Locations of new R&D sites
Source: Booz Allen Hamilton and INSEAD, 2006.
We may list several aspects that used to justify the geographic centralization of the R&D
function at a single unit.
To Klaus Brockhoff (1998), these reasons are related to: attempts to achieve economies of
scale, by keeping several researchers at a single site; the high cost of communication in case
the R&D function is decentralized into several units; ease of managing the R&D portfolio
3
4. and controlling proprietary information; and partnerships with public research institutes at the
company’s home country that provide new research perspectives without the need for
decentralization.
Ove Granstrand and colleagues (1992) also consider control of proprietary information to
be one of the main reasons for keeping R&D centralized at the company’s home country. To
the authors, such centralization minimizes the risk of competitors gaining access to the
research conducted. For these reasons and others, the authors propose that, should it be
necessary, decentralization should always be limited to a handful of countries. The authors
stress that keeping the R&D unit at a single site minimizes the costs of coordination and
control. John Cantwell (1992) also considers cost to be the main factor driving centralization.
Nevertheless, centralization may carry a higher total cost; in addition to coordination and
control costs, several others that may alter total site cost must be considered, such as the cost
of human resources, which may be lower in countries where hourly labor is less expensive.
Arnoud De Meyer (1992) summarizes that the main disadvantage to decentralization of
the R&D function is posed by communication difficulties, which hamper communication and
reduce development speed, eventually leading to decreased productivity. De Meyer believes
personal contact is the best form of communication in the R&D setting.
Figure 3 below outlines the advantages of a centralized R&D model, showing the
necessary inputs for R&D site activity, the reasons that lead to centralization, and the results
fed back to site clients.
4
5. Figure 3 – Advantages of the centralized R&D model
Source: the authors, after Vasconcellos et al, 2009.
Many of the aspects that once justified centralization are no longer considered valid
due to recent technological advances, particularly in communication technology.
The main advantages of geographic R&D decentralization are presented below.
To Klaus Brockhoff (1998), internationalization of sales and manufacturing – which,
depending on country, has specific economic and legal aspects – is itself one of the reasons
that lead to decentralization. Brockhoff considers the high level of researchers’ technical
qualifications to be the main factor of R&D. A company’s home country will not always have
the amount of highly qualified professionals necessary for research activities, and this
becomes one of the factors that justify decentralization.
Some corporate characteristics may actually demand decentralization. To Granstrand
et al. (1992), these characteristics may be divided into two categories: demand and resources.
Analysis of these factors, beginning with demand, shows that companies generally start their
external activities by means of export, only to later – with an increase in the demand of client
countries, following the classical process of internationalization – move in with a production
unit, which, depending on the type of product manufactured, may require the implementation
of an R&D site.
5
6. Foodstuffs are a good example: due to local customs, food products are more
commonly produced specifically for each location, even though major companies are
increasingly trying to keep global products in their portfolios. One way to assess this,
according to Granstrand et al. (1992), is by measuring the volume of customization requests
made by each country to the centralized R&D site. A high volume may justify the
establishment of a specific R&D center to meet local demand.
Another aspect concerns resources, which are truly a decisive factor to R&D
decentralization, particularly when key resources are located in another country. De Meyer
(1992) considers low cost of human resources combined with proximity to clients and access
to technology to be decisive factors for decentralization. De Meyer sees taking advantage of
lower overseas costs as a competitive advantage.
In the case of acquisition activity, when the acquired company has an R&D site
abroad, factor to be taken into account is technological complementary. If the acquired R&D
site brings complementary innovation, it should be kept at least until these innovations are
absorbed by the main R&D center; often, the cost of keeping a separate site is lower than the
cost of unifying a secondary site with the main center. Lars Hakanson (1992) believes that a
large number of R&D sites are located abroad because of corporate acquisitions, particularly
when companies are acquired for product diversification purposes; in this case, Hakanson
considers it the rule that existing R&D sites be kept and expanded.
To John H. Dunning (1994), reasons for decentralization are decisive when they truly
provide clear competitive advantages, which he divides, in a summarized manner, into four
aspects: when they bring improvements in processes, products, or materials; research for raw
materials or products; cost rationalization; and when the company obtains innovative activity
abroad.
Internationalization of R&D is becoming more a strategic issue for Frost (2001) and
Frost, Birkinshaw & Ensign, (2002). Multinational companies have R&D facilities outside
their home country to explore new knowledge or exploit of their current abilities. According
to Steven D. Eppinger et al. (2006), the new technology used to product development process
is fully digitally and completely networked, what make easier to decentralize.
R&D activities can also be influenced by the governments of each country.
Companies may set up R&D sites abroad through fiscal incentives, but the main factor occurs
when the government of a country enacts legislation that makes investing in research a
condition for the company to operate in the country’s market.
The decentralized model requires greater control of R&D activity, particularly
because such activity is usually conducted by virtual teams scattered throughout several sites
in different countries. A study on R&D decentralization at 3M (Vasconcellos et al., 2009)
showed why the Brazilian subsidiary of 3M was authorized by headquarters to adopt the
highest possible level of decentralization, which led to the development of new technologies.
The study focused on two aspects: how to establish the degree of technological complexity of
a decentralized R&D unit and how to identify which strategic technologies should be
assigned to each decentralized R&D unit. This study provided evidence of the need to
6
7. implement a worldwide R&D coordination network to minimize the disadvantages of
decentralization.
Figure 4 summarizes our review of the literature on the advantages of R&D
decentralization.
Figure 4 – Advantages of a decentralized R&D model
Source: the authors, after Vasconcellos et al., 2009.
International R&D decentralization is made up of four components that must be
analyzed and structured, not only to aid the process that underlies the decision of whether to
decentralize, but also to define which activities are to be undertaken and how the R&D site is
to be managed.
7
8. Figure 5 – Components of decentralization
Source: the authors
Geography – Establishment of a new R&D site abroad. This component involves
several interrelated decisions on the convenience of decentralization and on country choice.
This chapter will discuss the method used to select the most appropriate country for the new
site.
Activities to be undertaken by the new unit – A decentralized R&D unit may
contribute in several ways: product improvement and customization based on technologies
developed by headquarters in a centralized manner, customer service, and sales support. In
some cases, however, decentralized units conduct research on new technologies, as does
Rhodia’s Brazilian R&D site, for instance. Some decentralized units coordinate worldwide
R&D projects and spearhead the company’s efforts in certain technologies.
Decision-making autonomy – This component involves definition of the degree of
autonomy afforded to the decentralized unit in terms of project portfolio approval, R&D
budget, hiring, acquisition of equipment, expansion of lab facilities, outsourcing of R&D
services, approval of technology partnerships and technology purchasing, etc.
Subordination – When a company’s R&D sites are located in several countries, these
sites may be subordinate to a single corporate Director of R&D or similar position
(centralized subordination) or they may be subordinate to the directors of distinct business
units (decentralized subordination). These business units, in turn, may be subordinate to
products, clients, countries, or regions. In some cases, subordination is manifold, and
constitutes the use of matrix structures.
The above components require integrated treatment, as there are countless interfaces
among them. Geographic decentralization, for instance, must be accompanied by a certain
degree of activity and authority decentralization if the new R&D site is to contribute towards
company results.
8
9. To Frédérique Sachwald (2008), “The traditional contrast between the forces of
centralization and the forces of dispersion of multinationals’ R&D does not account for the
changes under way. Companies progressively find ways of combining the need for some
centralization of research with access to heterogeneous scientific and technological resources
throughout the world”
This chapter will focus on the choice of country. Several aspects must be considered
in this process, from factors that will guide center strategy, such as frequency and complexity
of contracts established in the country, to complementary factors such as incentives provided
and demands made by each country’s government. These factors should influence R&D
performance criteria: costs, deadlines, and quality of solutions.
After analyzing the work of several authors on the factors that allow or impede
decentralization of R&D units, we collated the main factors that must be assessed for the
strategic definition and determination of the most adequate location for international sites.
Figure 6 shows a conceptual model meant to aid the decision of choosing which
country is most appropriate to host a new decentralized unit. Three critical dimensions –
market, R&D inputs, and country profile – were added to the components in Figure 5. For
each dimension, we note the influence factors that should be analyzed.
9
10. Figure 6 – Conceptual model for country choice
Source: the authors
10
11. On the basis of Figure 6, we constructed a table to assess the importance of each factor for
country selection. Mr. Rezelman of SAP Labs Brazil was the respondent. The WEIGHT of
each statement to the decision-making process and company strategy was graded on a scale
of 1 (little importance) to 3 (very important). Each statement was then graded on a scale of 1
(low) to 10 (high) for each factor in each country, which will determine the comparison
between them. The respondent was instructed to leave his answer blank whenever a factor
was not applicable or when there was not enough information to support an answer.
Table 1 – Instrument designed for collection of information to support country choice
Country 1 Country 2
Degree of Degree of Country Country
Statements concerning the decision to establish an R&D unit Factor agreement agreement 1 2
in Country 1 rather than Country 2 Weight with with Total Total
statement statement
Country
Market Current market size for SAP products
Potential market growth for SAP products
Current and potential volume of development and
customization activities in the country
R&D Universities and institutions providing graduates and
Inputs conducting research in areas relevant to company
Conditions for obtaining equipment and material for R&D
activities at adequate quality and cost levels in the country
Competence and size of relevant talent pool
Cost of specialists at different career levels
Country
profile Political and economic stability
Proximity/ease of interaction with other countries and
headquarters
Availability and quality of infrastructure: Internet/broadband,
international airport.
Country provides adequate intellectual property and
confidentiality protection
Country has adequate mechanisms and legislation in place to
ensure fair competition and contract enforcement
The country has potential for innovation in areas that are of
strategic importance for SAP
SAP competitors have similar units in the country
Country’s cultural traits (e.g. language) will facilitate
implementation of new unit and cooperation with other R&D
units
Fiscal incentives for implementation of R&D sites in country
Government encourages that investment is made in local
R&D
Total
11
12. When SAP decided to set up a R&D centre in South America it considered a host of
countries, but in particular, Mexico, Chile, Argentina and Brazil. In the initial round the
above model was used, but with a focus on availability of talent (technical and language
skills), relative costs (labor and infrastructure), market size and potential, the country’s
economic and political stability and the opportunity for growth. The final round was between
Brazil and Argentina.
4. The company
SAP is a multinational business software company. As of December 2008, it employed
51,536 people around the world. Founded in 1972 as Systems Applications and Products in
Data Processing in Germany, it has since become the largest maker of enterprise resource
planning software in the world. It began operations in Brazil in 1995, through a marketing
and sales subsidiary.
The company began innovating with a development of enterprise software. After the
launch of SAP R/2, which allowed integrated management of corporate processes
(particularly in the manufacturing area), the firm expanded its product and solution portfolio,
with SAP R/3, CRM (Customer Relationship Management), and SCM (Supply Chain
Management), currently reaching SAP Business Suite, in addition to products for the mid-
size and small business segments and others.
Figure 7: Transformation from a One-Product Company to a Multi-Product Company
Source: SAP AG, 2009
SAP stands out not only in software development, but also for implementing innovation
in its corporate processes, which undergo continuous change and evolution. In 1984, the
company began its internationalization process, expanding its sales of SAP R/2 into other
countries besides Germany. In 2008, it reached more than 12 million users in over 120
countries, with earnings of 11,567 million Euros (Figure 8).
12
13. Figure 8: Evolution of SAP, 1999–2008
Source: SAP AG. Revenue and income in million Euros, 2009.
One may say that corporate management systems first appeared in the 1970s with
MRP (Material Requirement Planning), and then evolved to ERP (Enterprise Resource
Planning), which involves not only the materials area, but practically every process in the
company. After ERP came CRM, SCM, BI, and other acronyms that have become popular
with companies in general, all with the same objective: informatization of corporate
processes.
A management system must adhere to company processes. Internationalization
therefore requires much more than translation into the local language: identification of
specific needs and specific processes and true localization using the language of the segment.
The localization process seeks to develop specific modules and processes requires by the
legislation or culture of each country. A good example is the tax law framework of Brazil,
which featured unique characteristics that must undergo localization before they can be
implemented as part of a management system.
One may note that the internationalization of a management system is far more
complex than the translation of software such as operating systems, word processors, etc. It
requires, for one, both country specific functional and legal requirements to be incorporated.
Many companies that developed management software were acquired by larger organizations
over time, and, as of 2007, more than 50% of market share was held by three major software
makers: SAP, Oracle, and Microsoft.
Multinational ERP software makers only reached Brazil in 1995, due to several factors,
including the country’s closed market policy, technology gap, and localization difficulties;
SAP was the first such company to enter the Brazilian market. Meanwhile, local companies –
Microsiga and Datasul, now subsidiaries of the Totvs group – developed specific systems to
meet the requirements imposed by Brazilian legislation. In 1995, SAP entered the Brazilian
market and began the product localization process. This process used to be conducted by
means of separate modules, but since 2003 has been built into the system “Core”, making it a
13
14. global product, allowing easier updates, and making new versions of localized packages
available simultaneously at worldwide launch dates.
5. R&D Strategy
In 2008, SAP invested 14.1% of its earnings towards R&D, slightly down from the
previous year, but an increase on 2006. SAP had 15,547 employees dedicated to research and
development activities as of 2008 – a 20% increase from the 2007 figure. The company’s
R&D function contributes significantly to its product portfolio and expansion of market
leadership, through the identification and development of emerging IT trends and the
generation of technological advances. The product areas and development labs work on new
software functions and versions, and the research sites explore opportunities that have not yet
been developed into any product on the market. These opportunities may become products 2
to 10 years in the future.
Figure 9: R&D Centers
Source: SAP AG, 2009
SAP’s business model for research and development is based on innovation through
collaborative research with more than 520 organizations: universities, partners, clients, and
product groups from within SAP itself. SAP R&D sites are divided into different categories:
specific research centers, development centers, service centers, and research laboratories
working in cooperation with other research institutes. As shown in Figure 9, these sites,
which may be in a single location or in different places, are present in 16 countries (Australia,
Brazil, Bulgaria, Canada, China, France, Germany, Hungary, India, Ireland, Israel, Japan,
Korea, South Africa, Switzerland, and the United States).
The main unit is located at SAP headquarters in Walldorf, Germany; 48% of the
company’s R&D employees work there, although this percentage has been decreasing in
recent years (2006: 52%; 2005: 57%), as the company has increased its overseas investment.
14
15. Twenty-five percent of employees are stationed at development units in China and India, and
the remaining 27% throughout the company’s other sites.
Figure 10: Example of an R&D structure of SAP AG
Source: SAP AG, 2009.
SAP follows a distributed development model, coordinated by so-called Business
Units that keep development teams in the R&D sites of several countries. As shown in Figure
10 and 11, this is a matrix model, which requires rigorous control to guarantee a standard
quality and efficiency so in the end an SAP product is produced, irrespective where it comes
from.
Figure 11: SAP AG team matrix
Source: SAP AG, 2008.
The decision to decentralize occurred in the early 1990s, as the company sought to come
closer to the market, its clients and partners, technology, better practices, and talent. A global
company must also be present in many locations in order to better understand the needs of the
15
16. market and those of its clients. In the same manner in which SAP keeps R&D centers in
Shanghai and Bangalore to support its activities in Asia and part of Europe, the São Leopoldo
center in Brazil supports its activities in Latin America.
6. SAP’s new R&D unit: the choice of country
After assessing all of SAP’s investments on R&D sites, the question remains: why did
the company set up a site in Brazil? We applied our proposed model to the SAP case, giving
particular consideration to the final dispute between Brazil and Argentina.
Brazil Argentina
Degree of Degree of Brazil Argentina
Statements concerning the decision to establish an R&D Factor agreement agreement Total Total
unit in Country 1 rather than Country 2 Weight with with
statement statement
Country
Market Current market size for SAP products 2 3 1 6 2
Potential market growth for SAP products 3 3 2 9 6
Current and potential volume of development and
customization activities in the country
2 3 1 6 2
R&D Universities and institutions providing graduates and
Inputs conducting research in areas relevant to company 3 3 3 9 9
Conditions for obtaining equipment and material for
R&D activities at adequate quality and cost levels in the
country 2 2 2 4 4
Competence and size of relevant talent pool 3 3 3 9 9
Cost of specialists at different career levels 3 3 2 9 6
Country
profile Political and economic stability 3 3 1 9 3
Proximity/ease of interaction with other countries and
headquarters 2 3 3 6 6
Availability and quality of infrastructure :
Internet/broadband, international airport, 2 2 3 4 6
Country provides adequate intellectual property and
confidentiality protection 1 3 1 3 1
Country has adequate mechanisms and legislation in
place to ensure fair competition and contract enforcement 2 2 2 4 4
The country has potential for innovation in areas that are
of strategic importance for SAP 2 2 2 4 4
SAP competitors have similar units in the country 1 2 1 2 1
Country’s cultural traits (e.g. language) will facilitate
implementation of new unit and cooperation with other
R&D units 3 2 3 6 9
Fiscal incentives for implementation of R&D sites in
country 2 2 2 4 4
Government encourages that investment is made in local
R&D 1 1 1 1 1
Total: 95 77
16
17. We identified the following as the main factors (weight grade 3) for SAP to establish
an R&D unit in a certain country:
Potential of market growth for SAP products.
Demand is one of the most important factors when choosing the location of a new
R&D site, because the country with the greatest demand may be better served by a nearby
site. This factor is directly tied to stability, as political and economic stability ensures demand
and return on investment.
Country’s political and economic stability.
Universities and institutions providing graduates and conducting research in
areas relevant to company.
The presence of universities and other institutions conducting research in areas that
interest the company is considered important, because research institutions allow the
establishment of partnerships and provide highly qualified professionals (which is also a
relevant factor).
Competence and size of relevant talent pool.
R&D sites require highly qualified professionals if research and development
activities are to provide satisfactory results. The country needs talent to justify investment.
This also justifies the fact that the major IT firms have R&D sites in India, as it is
considered a major hub of highly qualified IT professionals.
Cost of specialists at different career levels.
To SAP, the cost of experts – lower than in Germany – is a factor that complements
the preceding one, because it makes site implementation feasible.
Country’s cultural traits (e.g. language) will facilitate implementation of new
unit.
These factors were confirmed in the survey conducted by Booz Allen Hamilton and
INSEAD (Doz et al., 2006), with variations in their percentage from country to country, but
nonetheless considered most relevant (see Figure 12).
17
18. Figure 12 – Factors influencing the selection of locations for implementation of R&D sites
Source: Booz Allen Hamilton and INSEAD, 2006.
The market potential factor was found to be most relevant in practically all countries,
except for India, Germany, and the UK, where it came second to quality of workers (which,
in turn, came second in Brazil, with 21%). As clearly shown, these factors were considered
relevant, and may be taken to be deciding factors for the decentralization process.
In 2007, Latin America accounted for 9% of SAP’s global earnings, and Brazil accounted
for 40% of this figure. Latin American earnings increased 12% as compared to the previous
year, in accordance with SAP’s worldwide goal of double-digit yearly growth. SAP
considered four Latin American countries in its initial analysis: Mexico, Chile, Argentina,
and Brazil.
In summary, Mexico was considered to have a great talent pool and adequate
infrastructure for implementation of an R&D site. The two main reasons that prevented
investment in the country were that due to its proximity to the North America site, it had
relative higher costs of talents and infrastructure. In analyzing Chile, SAP noticed extensive
government incentives for the IT sector. Due to a small SAP product client base and limited
talent availability, which would only be adequate for the initial stages of implementation but
not for the desired growth plan, the company did not believe Chile would be the best option.
In light of these analyses, the choice was down to Brazil or Argentina. A comparison of
Brazil and Argentina using the methods proposed in this chapter yielded 95 points for Brazil
and 77 for Argentina.
The score was the same for both countries on several factors, such as specialist
competence, infrastructure, and others, but some differences favored Brazil as the chosen site.
18
19. Argentina fared better with regard to the following factors:
1) Availability of Internet, airport, personal security, and contract enforcement
infrastructure in the country. For this factor, we may stress the ease of obtaining flights
from Argentina to Germany and to other countries in which the center must operate.
2) Country’s cultural traits (e.g. language) will facilitate work of new unit in the
region. As Argentina is a Spanish-speaking country, choosing it would greatly facilitate
communication with other countries in Latin America. This is a disadvantage for Brazil,
where Portuguese, not Spanish, is the national language.
Brazil stood out on the following factors:
1) Local market:
– The size of the current market for SAP products is greater in Brazil;
– Brazil offers greater growth potential for SAP products, and is part of BRIC (group of
emerging economies that includes Brazil, Russia, India, and China);
– The volume of development and customization activity requested by Brazil is greater
than that of Argentina, which is justified by the size of the existing client base. Overall the
revenue base is considerably larger than Argentina as well.
2) R&D inputs:
– The cost of specialists is lower in Sao Leopoldo compared to Buenos Aires. In addition in
Brazil talents are split over several cities, whereas in Argentina they were mainly
concentrated around Buenos Aires, which impacts flexibility.
3) Country profile:
– The political and economic stability of Brazil was also considered to be an advantage over
Argentina, which, in addition to political instability, has higher inflation than Brazil, and
therefore poses a greater degree of investment risk;
– The country provides adequate protection for intellectual property rights and
confidentiality. Brazil is ranked on par with Canada by the US IP watchdog;
– SAP competitors have similar units in the country. Google, Microsoft, HP, Dell operate
similar centers in Brazil though with a smaller scope. It was shown that it is also the most
likely location for future expansion.
In light of these factors, Brazil was chosen in 2006 to host the newest SAP research and
development site.
7. SAP LABS at São Leopoldo
On June 5, 2006, SAP opened its Brazilian R&D site at the Unisinos (Universidade do
Vale do Rio dos Sinos) campus in São Leopoldo, Rio Grande do Sul. The city of São
Leopoldo was chosen because of the University and the linked Technology Park
(TechnoSinos), which is one of the oldest in the state. The campus is also very green,
providing the right environment for creativity and innovation. Overall, the cost is 30% to
40% lower there than in São Paulo. São Leopoldo is also safer, has the necessary
19
20. infrastructure available, and has excellent IT talent in the area (including Porto Alegre and
Novo Hamburgo).
The main focus of this center is development and support of SAP products for the
Americas, Italy, Spain, and Portugal. The center customizes products, develops new ones
(such as electronic invoicing technology and specific products for certain segments such as
the oil industry), conducts product testing and conducts product research in association with
other SAP centers. SAP’s development is set-up as a truly global organization in that it also
distributes core development to its Labs Network, to align with availability of talents, know-
how, innovation and costs. Most other multinational software companies maintain a central
core development site in their home country and only distribute localization and translation to
foreign sites.
The São Leopoldo Center is part of the SAP R&D network: it receives requests from
business units to take part in development and research of SAP products. The percentage of
development activities exceeds that of research, although the research portion of site
activities is expected to expand over time. This factor also features as a market strategy in the
survey conducted by Booz Allen Hamilton and INSEAD (2006).
Figure 13 – Activities of Home-Base and Foreign R&D Sites
Source: Booz Allen Hamilton and INSEAD, 2006.
In foreign R&D sites, 55% of activities were found to focus on development rather than
research. The São Leopoldo Center has a highly productive and scalable structure,
comparable to that of sites in India and China, which is a pre-requisite for its future growth
potential. As of Sept 2009, the site had 258 employees, a figure that is expected to grow to
370 by late 2010 and 800 in 2013. Ninety-seven percent (97%) of employees are from the
São Leopoldo/Porto Alegre area, which allows achievement of another of the center’s goals –
the development of local talent. The remaining 3% are from other SAP locations to help with
the set-up, knowledge transfer and implementation of SAP best practices.
20
21. As part of SAP’s sustainability initiative, the São Leopoldo Center has since June 2009
moved into a custom build, modern campus; the building was constructed and will be run
according to the US Green Building Council’s LEED Gold “good environmental practices”
certification. With an environment free of toxic substances and materials, ample availability
of natural light, fresh air, and a focus on the well being of its employees, this site may even
exceed the results obtained in China and India, helping Brazil reveal major talents in the IT
field and, through these talents, perhaps influence other areas of knowledge.
Each company must assess the most adequate R&D structure model (centralized or
decentralized) according to its characteristics. Our suggested methodology proved useful for
choosing the most adequate country to host an R&D site. Correct strategic decisions bring
returns on investment within estimated timeframes, and allow corporate growth. Brazil
currently has several characteristics that are very relevant to receiving R&D investment,
particularly economic stability and sustained growth figures, which allows other companies
to invest in the country.
We acknowledge the need for further studies if our conclusions are to be generalized, as
the case study method does not allow this. On the other hand, analysis of the SAP case
demonstrated the potential of our model for use in other organizations if necessary
adjustments are made.
8. References
Asakawa, K., & Som, A. (2008). Internationalization of R&D in China and India: Conventional
wisdom versus reality. Asia Pacific Journal of Management, 25(3), 375-394.
Belderbos, R., Lykogianni, E., & Veugelers, R. (2008). Strategic R&D Location in European
Manufacturing Industries. Review of World Economics, 144(2), 183-206.
Brockhoff, K. (1998). Internationalization of Research and Development. Springer.
Cantwell, J. (1992). The Internationalisation of Technological Activity and its Implications for
Competitiveness. In O. Granstrand, L. Hakanson, & S. Sjolander (Eds.), Technology
Management and International Business: Internationalization of R&D and Technology,
Chichester: Wiley.
Coeurderoy, R., & Murray, G. (2008). Regulatory environments and the location decision:
evidence from the early foreign market entries of new-technology-based firms. Journal of
International Business Studies: Part Special Issue: Institutions and International
Business, 39(4), 670-687.
Dachs, B., Ebersberger, B., & Lööf, H. (2008). The innovative performance of foreign-owned
enterprises in small open economies. Journal of Technology Transfer, 33(4), 393-406.
De Meyer, A. (1992). Management of International R&D Operations. In O. Granstrand, L.
Hakanson, & S. Sjolander (Eds.), Technology Management and International Business:
Internationalization of R&D and Technology. Chichester: Wiley.
Doz, Y., Wilson, K., Veldhoen, S., & Altman, G. (2006). Innovation: Is Global the Way Forward.
A joint study by Booz Allen Hamilton and INSEAD. Booz Allen Hamilton & INSEAD.
Dubois, A., & Gadde, L.E. (2002). Systematic combining: An abductive approach to case research.
Journal of Business Research, 55(7), 553–560.
Dunning, J. H. (1994, January). Multinational Enterprise and the Globalization of Innovatory
Capacity. Research Policy, 23(1), 67–88.
Easton, G. (1995). Case research as a methodology for industrial networks: A realist apologia.
Proceedings of the 11th IMP Conference, Manchester, 368–391.
21
22. Edler, J. (2008). Creative internationalization: widening the perspectives on analysis and policy
regarding international R&D activities. Journal of Technology Transfer, 33(4), 337-352.
Edler, J., & Polt, W. (2008). International industrial R&D--policy challenges: introduction to the
special issue. Journal of Technology Transfer, 33(4), 331-336.
Eisenhardt, K. M. (1989). Building theory from case study research. Academy of Management
Review, 14(4), 532–550.
Eppinger, S. D., Chitkara, A. R. (2006). The New Practice of Global Product Development. MIT
Sloan Management Review, 47(4), 22-30.
Fisch, J. (2008). Investment in new foreign subsidiaries under receding perception of
uncertainty. Journal of International Business Studies, 39(3), 370-386.
Francesca Sanna-Randaccio, & Reinhilde Veugelers. (2007). Multinational knowledge spillovers
with decentralised R&D: a game-theoretic approach. Journal of International Business
Studies, 38(1), 47-63.
Frost, T. S. 2001. The geographic sources of foreign subsidiaries' innovations. Strategic
Management Journal, 22(2): 101-123.
Frost, T. S., Birkinshaw, J. M., & Ensign, P. C. 2002. Centers of excellence in multinational
corporations. Strategic Management Journal, 23(11): 997-1018.
Gassmann, Oliver and Von Zedtwitz, Maximilian (2003), Trends and Determinants of Managing
Virtual R&D Teams. R&D Management, Vol. 33, pp. 243-262, June 2003.
Granstrand, O., Hakanson, L., & Sjolander, S.(1992). Technology Management and International
Business: Internationalization of R&D and Technology. Chichester: Wiley.
Guimón, J.. (2009). Government strategies to attract R&D-intensive FDI. Journal of Technology
Transfer, 34(4), 364-379.
Gulbrandsen, M., & Godoe, H.. (2008). "We really don't want to move, but...": identity and
strategy in the internationalization of industrial R&D. Journal of Technology
Transfer, 33(4), 379-392.
Hakanson, L. (1992). Locational Determinants of Foreign R&D in Swedish Multinationals. In O.
Granstrand, L. Hakanson, & S. Sjolander (Eds.), Technology Management and International
Business: Internationalization of R&D and Technology, Chichester: Wiley.
Hemmert, M.. (2003). International Organization of R&D and Technology Acquisition
Performance of High-Tech Business Units1. Management International Review, 43(4), 361-
382.
Khalil, T. M., Lefebvre, L. A., & McSpadden, R. (2001). Mason Management of Technology: The
Key to Prosperity in the Third Millennium. In Selected Papers from the Ninth International
Conference on Management of Technology. Emerald Group Publishing.
M. von Zedtwitz and O. Gassmann, “Market Versus Technology Drive in R&D
Internationalization: Four Different atterns of Managing Research and Development,” Research
Policy 31, no. 4 (2002): 569-588.
McCutcheon, D., Meredith, J. (1993). Conducting case study research in operations management.
Journal of Operations Management, 11(3), 239–256.
Miles, M.B., Huberman, A.M. (1994). Qualitative Data Analysis: Grounded Theory Procedures
and Techniques. London: Sage Publications.
Rama, R.. (2008). Foreign investment innovation: a review of selected policies. Journal of
Technology Transfer, 33(4), 353-363.
Sachwald, F.. (2008). Location choices within global innovation networks: the case of
Europe. Journal of Technology Transfer, 33(4), 364-378.
SAP. (2008). SAP RESEARCH REPORT 2007/2008. Retrieved September 1, 2008, from
http://www.sap.com/about/company/research/pdf/SAP_Research_Report_2007-2008.pdf
SAP. (2007). SAP Annual Report 2007. Retrieved September 1, 2008, from
http://www.sap.com/about/investor/reports/annualreport/2007/pdf/0417_SAP_GeBe07_USGA
AP_GB.pdf
22
23. Vasudeva, G., & Sonderegger, P. (2007). R&D internationalization: Building organizational
capabilities to balance exploration and exploitation. In Tallman, S. (Ed.). A New Generation in
International Strategic Management. Northampton, MA: Edward Elgar Publishing.
Vasconcellos, E., Silva, L., Guedes, L. & Vasconcellos, L. (2009, April). Internationalization of
R&D at 3M focusing on Wind Energy Business – Conceptual Model and Application. In
IAMOT, International Conference on Technology Management, Orlando, FL, USA.
Vasconcellos, E., & Vilarino, V. (2008). Centralização x Descentralização Internacional de P&D.
Unpublished manuscript, University of São Paulo School of Economics and Business
Administration.
Yin, R., (1994). Case Study Research. Beverly Hills, CA: Sage Publications.
23