Opinion
Computing Applications

International Perspectives: Software Strategies in Developing Countries

Posted
  1. Introduction
  2. Are Exports 24-Karat or Fool's Gold?
  3. Roadblocks
  4. Opportunities
  5. A Third-World Microsoft?
  6. Small Fish in a Small Pool
  7. Straddling the Intersections
  8. Understanding the Winners
  9. References
  10. Author
  11. Footnotes
  12. Figures

Globalization touches everything these days, and software is no exception. It’s big business in developing countries (DCs),1 from the headline-grabbing software factories of Bangalore to the back-street database designers of Nairobi. And no wonder: many of these countries have cheap, talented labor that finds an easy route into a business that is labor-intensive, has relatively low entry barriers, and has few economies of scale. Computer science graduates need arm themselves with just a PC and a couple of user contacts to become part of the local information economy. Add a modem and they are global “infopreneurs.”

But this is also a multifaceted business. Which software targets should DCs be aiming for? Experience shows five strategic positions that can be taken by DC software enterprises, shown in figure.

Back to Top

Are Exports 24-Karat or Fool’s Gold?

At first sight, the export-oriented strategy represented by positions A and B in the figure seems highly attractive. As noted, the entry barriers appear low and the rewards appear high. Extrapolation from the patchwork of incomplete and sometimes uncertain figures suggests DCs exported some (U.S.) $3 billion of software in 1998–1999. However, look behind the mask to understand the reality of these exports.

The export image projected is one of virtual development, in which clients sitting in the West interact with software professionals developing packages overseas. The reality, though, is somewhat different with most software exports having the following profiles.

A skewed output profile. Software packages represent a small proportion of total DC exports. Instead, producers provide software services, which account for more than 70%, 80% and 95% of Russian, Philippine and Indian exports, respectively [2, 6]. The software success stories have therefore been overwhelmingly in position A rather than B.

A skewed location profile. Large amounts of development work take place at the client’s site by having DC software developers fly over to work with the client. For some, the skews are even greater. Pakistan exports around (U.S.) $15 million of software per year [10]. Its main software exporters now manage their software export operations from offices in the U.S. and employ U.S.-trained Pakistani programmers, raising the question of what it means to “export” software and to be a “DC” company.

A skewed skills profile. Most work undertaken by DC developers is relatively low-skill software construction and testing, leaving the high-skill tasks of analysis and design residing in Western hands. Exporters from India—the Third-World’s software export giant—do take on a limited number of cradle-to-grave contracts, but 80% or more of earnings come from grunt work [6].

Despite these skews, the high profile of India’s success makes many DCs wish to follow in its footsteps. Those who do, however, find roadblocks along their path.

Back to Top

Roadblocks

The infrastructure roadblock. The software export trade increasingly demands a sizeable installed computer base, reliable and pervasive telecommunications links both domestically and internationally, and reliable electricity supply. Rapid strides are being made in overcoming this deficiency, as the global spread of telecommunications attests. However, inequalities persist and some work must always be done at the client’s site. Telecommunications technology is thus slowly modifying this trade but has yet to revolutionize it.

The FUD roadblock. Many Western clients still express “fear, uncertainty, and doubt” about DC contractors and their business environment [8]. Ask Western business managers to conjure an image of the Third World, and they will tell the same story as the mass media—poverty, famine, war, and corruption. Such stereotypes are hardly fertile ground in which to sow the seeds of software export, especially when combined with the toxin of outsourcing disaster stories. For example, Hewlett-Packard was taken to court and (worse) paraded on 60 Minutes over supposed exploitation of low-cost Indian contractors; insurance firm AIG outsourced an entire division’s programming to Indian staff who were not up to the job. Thus, despite the massive growth in trade, sourcing software from DCs remains anathema to large swathes of Western business. Only 16% of U.S. corporations see this as a significant strategy to meet their software needs [7].

The late-comer roadblock. A number of DCs, such as India and Singapore, arrived on the export scene many years ago. So, too, countries on the European periphery—such as Ireland, Israel, and Hungary—have been low-cost software export bases since the early 1990s. These countries have already built up contacts, policies, infrastructure, working methods, and track records. As a result, the more established players threaten to consolidate their position while squeezing out late-comers [5, 11].

Back to Top

Opportunities

In addition to first-mover consolidation of existing export markets, three further opportunities arise within position A.

For one, first-movers have been adept at service market diversification. Indian software exports have registered real-term annual growth of 40% or more—nearly (U.S.) $2 billion in 1998—by successfully surfing the waves of demand from downsizing to Y2K and on to conversion for the new Euro currency [11].

First-movers have also become concerned to move up the value chain. They aim to do this through industry-sector specialization that will both deepen trust relations with their clients and deepen their expertise. Add in strong software increases in the U.S. H1B visa quotas used for foreign workers and—partly countervailing the squeeze-out of late-comers—some limited opportunity gaps may be opening up for “second-comers” to the software export business. Recent exports from countries such as Sri Lanka and Mauritius support the existence of such gaps.

Inter-DC markets are also growing. Africa, Asia, and the Middle East are the target for about 5% of India’s software exports, while Korea and Malaysia are outsourcing to countries like the Philippines, China, and Vietnam.

Of course, the realities of this strategy must be faced since position A is typically an export enclave in which skills and technology fail to trickle down to the domestic market. Staff who work on export projects are far more likely to receive Green Cards and move on rather than go back to domestic market work. India loses around 15% of its software workers every year, largely to the U.S. [5]. Some firms hemorrhage staff so fast that they “run in order to stand still,” undermining attempts to move up the value chain.

Even money can become partly enclaved. Much of that (U.S.) $3 billion leaves DCs to pay for travel and living allowances of the developers who work at the client site, marketing expenses, information and communication technology imports used for DC-based contract components, and profit repatriation by the many multinationals involved in this trade [9]. Net earnings are thus far less than gross, knocking the headline figure to less than (U.S.) $1.5 billion.


Some firms hemorrhage staff so fast that they “run in order to stand still,” undermining attempts to move up the value chain.


Opportunities must finally be set alongside opportunity costs. Putting your brightest software stars to work on applications that boost the growth of foreign firms and foreign economies incurs a large opportunity cost when applications to meet the many pressing domestic needs are consequently sidelined.

Back to Top

A Third-World Microsoft?

Position A is a successful strategy adopted by a few DCs, but not one that represents a panacea for all. Instead, then, should some software enterprises in DCs be aiming for the production of domestic software packages, following option C in the figure? In the general applications market of word processing software, spreadsheets, databases, operating systems, and the like, they certainly should not.

Imported packages, either legal or pirated, have that market (shrink-)wrapped up, and entry barriers for local firms remain formidable [3, 4].

Their low labor cost advantages are quickly eroded by packages’ high development costs. High rates of piracy squeeze an already small domestic market. Many DC consumers also favor foreign rather than local software, irrespective of price, quality, and features.

So strong are these cost, piracy, and foreign-preference roadblocks that, in sum, the Microsoft of the 21st century is not currently incubating in the technology parks of India or in Malaysia’s multimedia corridor.

Back to Top

Small Fish in a Small Pool

The vast majority of DC software firms sit in the position-D market segment, largely because it is by far the easiest for them to enter.

This can be a good starting point for progress into, say, exports. A sizeable and demanding domestic market can be the springboard from which to launch into exports by providing a base of relevant skills, experience, user feedback on products, and track record. Second, a sizeable domestic market will draw large numbers of IT multinationals into collaborative relationships with local partners in order to serve that market. As these relationships deepen, an export component often emerges.

Unfortunately, the domestic software market in many DCs cannot yet be described as either sizeable or demanding. In Western countries, software represents more than 55% of the IT market; in countries like China and Indonesia, the figure is less than 20% and there is a technology lag of some years behind the leading edge [12]. This creates a domestic enclave separated from the global market.


So strong are these cost, piracy, and foreign-preference roadblocks that, the Microsoft of the 21st century is not currently incubating in the technology parks of India or in Malaysia’s multimedia corridor.


Because of this domestic market roadblock, most position-D firms therefore remain position-D firms: small fish in a small pool that do not grow and which get very limited exposure to new ideas and new technologies. For them, position D is more a survival strategy than a development strategy.

Back to Top

Straddling the Intersections

Alongside strategy A, position E represents the other main success story of DC software, with a theme of specialization for niche markets that include:

  • Sectoral niches like banking, insurance, health administration, hotel management, mining and forestry;
  • Application niches such as Web browser add-ons or text retrieval utilities; and
  • Linguistic niches for regional languages like Spanish, Swahili or Chinese.

There is no single route for success here, but Chilean firms epitomize one major pathway. Typically they began by providing one client with Spanish interface software custom-built to meet particular local needs. They then created a “semi-package”: a set of menu or window interfaces used as a marketing or development platform for further customization initially in the local market and then, as opportunities were perceived, in the Spanish-speaking markets of Latin America and even Europe. The result has been exports of nearly (U.S.) $100 million [1].

Alternative routes include those of Israeli firms that have spotted export opportunities for domestic market packages, and even those of a few enclave-busting Indian firms that have repatriated ideas gained from service export contracts and incorporated them into customizable packages for the local market.

With its initial roots mainly in the domestic market, position E is under threat as that market comes to the attention of Western software producers. They have started to localize their packages—even some for niche markets—and to set up services subsidiaries. Nevertheless, through their local contacts and local understanding, DC producers retain a knowledge advantage that some can put to good use.

Generic tactics and strategy are discussed next, but successful “straddlers” also specifically depend on two aspects of market integration that mean they largely avoid the enclave and opportunity-cost shortcomings suffered by position A. First, an entrepreneurial recognition of opportunities for synergy and leverage is spotting a local niche that will also sell overseas or developing a package that will help sell customization services. Second, demanding local customers, typically a set of regionally or globally competitive businesses that provide a channel from one market segment to the next, is needed.

Of course there are roadblocks here too (shared with other strategic positions). As well as domestic market demand and integration constraints, there is a skills roadblock, with simple coders making up 85% or more of all software personnel [3, 5]. There is also a finance roadblock caused by conservative attitudes and lack of venture capital.

Back to Top

Understanding the Winners

What is it that has differentiated those who have succeeded from those who have not in DCs? Some reasons have already been described, but we can also identify more generic coordinated efforts made in three main domains by the winners.

1. Enterprise Tactics. Little is known about the management tactics and critical success factors that underpin software enterprises in DCs. In addition, successful software enterprises require effective mechanisms for information and knowledge transfer that rely on three forms of both formal and informal networking: internal networking such as meetings and team working; peer networking between local software firms, typically based either on the high levels of staff turnover for which the IT industry is renowned or on informal meetings; other external networking between the enterprise; and foreign markets and innovators, potential and actual clients, and government officials. The U.S.-based diaspora from India, Ireland, and Israel has been fundamental to their external networking and, hence, to their software export success, a reminder that brain drain has its benefits as well as costs.

In all this, we see why software firms have tended to cluster. Not only are there collective efficiencies in the supply of infrastructure, people, and money, but proximity also oils the wheels of informal knowledge exchange.

What we do know suggests that tactics in relatively open developing economies overlap significantly with those in the West. Successful development of software enterprises has been based on factors such as identification of demand-growth markets and synergies, cost or service innovation, and good marketing. Enterprises must also obtain sufficient access: to both investment and working capital; to programming, analysis, and management skills; and to information technology.

2. National Strategy. Enterprise tactics alone are not enough to produce critical success factors, partly because firms face the various roadblocks that they alone cannot overcome. Higher authorities must therefore become involved. America built its IT industry on government money pumped in during critical early growth years in the 1940s, 1950s, and 1960s. Similarly, government interventions to overcome roadblocks in DCs have underpinned the success stories.

This is not the big brother/nanny state model that many countries adopted in the 1960s and 1970s, which would threaten to smother an otherwise lively industry with red tape. It is, instead, the model of government as industry promoter. But what does it mean in practice for DC governments to become “promotional states” in support of local software enterprises? It means they act along a range of fronts [4, 9].

Finance. Successful governments act to stimulate the supply of working and venture capital to software firms, by enabling private sector and overseas funding. Countries like Israel and Taiwan have used a raft of tax breaks, marketing subsidies, grants, loans, legislative updates, and extermination of red tape in an effort to achieve this.

Education and training. The state is likely to remain the prime source of fundamental capabilities relevant to software industry development. The best also work with other providers—including private trainers—to target the analytical and managerial skills that the developing world so crucially lacks. Use of international accreditation and certification schemes for skills has helped to address commonplace training quality deficits. Ireland’s success in reinventing itself from poor, rural backwater to Celtic Cyber-Tiger has rested heavily on intensive investment in people, with half the population going into tertiary education and considerable emphasis on IT courses.

Research and development. Most publicly funded R&D systems need a radical overhaul, of which increased involvement of the private sector will form a centerpiece. Nevertheless, in some cases R&D has been a bulwark against the brain drain and against the growing concentration of innovation within multinationals. Best investments are focused on customization to meet local needs and on the commercialization of existing capabilities (via “straddling”), not on wasteful basic research inspired by an imported Western agenda. Israel has led the way here, with government-subsidized multimedia projects that have spun off into everything from games to business and home applications, and with military-funded developments in signal processing and encryption emerging in a variety of Internet communication and security packages.

Intellectual property rights. Piracy has more to recommend it as a strategy for DCs than is often admitted. It speeds diffusion of the local IT base, creates a broad foundation for learning through “reverse functional engineering,” and saves huge wads of foreign exchange. However, the maturation of a software industry goes hand-in-hand with a legal framework of IP laws and enforcement that only the state can provide. Such a framework is also a sine qua non for serious foreign investment. Microsoft, for example, only agreed to set up a software production facility in Egypt on condition of tougher government legislation against piracy.

Infrastructure. From Singapore’s intelligent island to South Africa’s telecenters, the state has played a vital part in the creation of a telecommunications infrastructure. Given public investment constraints, however, some infusion of private and foreign capital has been essential to this process. Governments in a variety of countries, such as Brazil, have also moved to create “software parks”—high-tech clusters that benefit from the economies of scale and informal networking noted.

3. National Vision. You can teach almost anyone the techniques of painting, but creation of a work of art requires an artist with vision who can apply those techniques in realization of their vision. So, too, strategies such as those described are all well and good, but they only emerge or sustain when there is a national vision for software. India now bestrides the software scene as a Third-World colossus thanks to a 30-year dream of software exports, shared by bureaucrats and industrialists alike but carried forward by some key “artists with vision.” Not all DCs can follow India’s path. However, they can reap the software rewards of high-tech jobs, capabilities, and income if they learn to combine successful tactics, strategy, and vision.

Back to Top

Back to Top

Back to Top

Back to Top

Figures

UF1 Figure. Strategic Positioning for Developing Country Software Enterprises

Back to top

    1. Baeza-Yates, R.A., Fuller, D.A., Pino, J.A. and Goodman, S.E. Computing in Chile. Commun. ACM, 38, 9 (Sept. 1995), 23–28.

    2. Computing Wired World. Computing global innovators series. (Jul. 10, 1997), 7.

    3. Correa, C. Strategies for software exports from developing countries. World Development 24, 1 (1996), 171–182.

    4. Grundey, M. and Heeks, R.B. Romania's hardware and software industry. IDPM development informatics, paper no. 2, University of Manchester, U.K., 1998.

    5. Heeks, R.B. India's Software Industry. Sage Publications, New Delhi, 1996.

    6. Heeks, R.B. The uneven profile of Indian software exports. IDPM development informatics paper no. 3, University of Manchester, U.K., 1998.

    7. Information Technology Association of America. IT Workforce Study. Arlington, Va, 1998.

    8. Kumar, B. Software men dubbed 'bad boys' by U.S. companies; Soc.culture.indian Usenet newsgroup, (Aug. 9, 1998).

    9. Mansell, R. and Wehn, U. Knowledge Societies: Information Technology for Sustainable Development. Oxford University Press, Oxford, 1998.

    10. Nicholson, M. Realism quashes euphoric projections. Financial Times (Aug. 12, 1997), 4.

    11. Taylor, P. New IT mantra attracts a host of devotees. Financial Times India IT Survey (Dec. 2, 1998), 1.

    12. World Trade Organization. Computer and Related Services. Geneva, Switzerland, 1998.

    1The term "developing country" as used here, encompasses not merely the nations of Africa, Asia, Latin America and the Caribbean, but also the transitional economies of Eastern Europe and other nations on the European economic periphery such as Turkey and Ireland.

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More