IT Services: Talent Communities

The talent shortage, customer demand for complex solutions and the size limits of Indian IT services firms will increase transaction costs to the point where large companies will start experiencing dis-economies of scale.
On the other hand plummeting technology costs and the internet will accrue economies of scale from the external environment to smaller firms.
While the size, at which a firm will start experiencing dis-economies of scale, depends on the industry and the external environment, it is moderated by two transaction cost-related factors: organization form and asset specificity.
Organizational form basically refers to the structure of the organization and how this structure helps or hinders coordination and communication. There are two basic types of organization: M-Form and U-Form. An M-form (multi-divisional) company is organized along product lines where complementary tasks are grouped together. Like a computer manufacturing company that is organized into the Server Division and the PC Division, with each division having its own functional units like sales and accounting.
In contrast, a U-form (unitary form) organization is process focused and is decomposed into functional units like a Sales Department and a Manufacturing Department, with each unit catering to all product lines.
Organizational forms affect coordination of two critical processes: processes that facilitate the assembling of parts or tasks and processes that create interchangeability or substitution of parts or tasks.
A simple trade-of emerges between two types of coordination and scale economies. The self-contained units of the M-form, allow easy assembly of complex products and services. The U-form organization, lends itself to interchangeability, standardization and economies of scale. However assembly of complex and fast changing products is difficult in U-form organizations.
Asset specificity refers to the extent to which a party is “tied in” in a two-way or multiple-way business relationship. For example, learning to speak English, the most universal language of the world, is a highly asset-unspecific investment, since your investment will likely to have equal returns (being able to communicate with others) across a variety of different settings. On the other hand, learning to speak Konkani, a language spoken in the west cost of India, could be highly asset-specific since your investment return is high with the few Konkani-speaking people, but almost zero otherwise.
A large organization that is able to create asset-specificity reduces diseconomies of scale for itself. For example the large population of Websphere and AS/400 programmers or SAP programmers is highly asset-specific to IBM and SAP, and reduces diseconomies of scale for them. In the information technology industry, however, the move towards open source and platform independent technologies is reducing asset specificity.
In a world where open source and standardization is reducing asset specificity and increasing cross-product substitution and compatibility, the gains from internal economies of scale are reducing. As customers demand increasingly sophisticated solutions and services which need the assembly of complex tasks and services, the need for M-form type organization will increases to a point where the large Indian IT firms will soon start experiencing diseconomies of scale. The immensely successful people centric growth model of Indian IT services firms may well start working against them, and they will soon find themselves at a point where the costs of organizing additional complex people intensive business transactions within the firm exceed the costs of carrying out the same transactions through the market or within another firm.
We can already see this starting to happen. Rising attrition rates and diseconomies of scale, are forcing some Indian IT companies to look at alternate organizational models. One of the pioneers is Satyam Computers, India’s seventh largest IT firm (2006-07 revenues of INR 6000 crores and 35,000 people). Satyam has been publicizing its very successful experiment with what it calls a ‘distributed leadership’ model, wherein business areas are owned and not merely ‘manned’ by a leader. The company went through a detailed organizational study, looked at all its processes and turned these into business units. So for example recruitment, both direct from the campus and lateral, was spun off into two business units. Different customer accounts, visa processing and even training and development were now treated as independent business units or companies run by virtual CEOs with complete authority over their domains.
The business units are benchmarked against the best players in their domain. For example the lateral hiring business started benchmarking itself against a placement organization. Already, the results of Satyam’s experiment are beginning to emerge. Its ADI or Associate Delight Index, which is a reflection of how people in the company feel, touched an all time high of 4.33 (on a scale of 5), and the company’s attrition rate has come down by 2 percent in the last one year.
Academic research as well, has been urging companies to adopt a more distributed organizational model in favor of the old hierarchical model.
In his book, “The Future of Work”, Thomas Malone makes the case that with the costs of communications plummeting and the value of the information communication increasing it is inevitable that organizations will decentralize more than ever by employing hybrid forms of loose hierarchies, democracies and free markets for the same organization.
In “Unbundling the Corporation”, John Hagel and Marc Singer, take the view that no matter how monolithic they may seem, most companies are really engaged in three kinds of businesses. One business attracts customers. Another develops products. The third oversees operations. Although organizationally intertwined, these businesses have conflicting characteristics. Historically, they have been bundled because the interaction costs incurred by separating them were too high. But as electronic networks drive down the costs of communicating and of exchanging data, we are on the verge of a worldwide reduction in interaction costs. Activities that companies have always believed were central to their businesses will suddenly be offered by new, specialized competitors that won’t have to make trade-offs. Ultimately, large traditional businesses will be forced to adopt ways to un-bundle themselves and then re-bundle into large infrastructure and customer-relationship businesses and small, nimble product innovation companies.
Is it Outsourcing?
Ever since 1988, when Eastman Kodak announced that it was outsourcing its information systems function to IBM, DEC and Businessland, outsourcing has steadily gained mainstream acceptance and it is very tempting to look at the unbundling of businesses as a form of outsourcing. But this may not be entirely correct. The core academic theories behind outsourcing are the theories of Comparative Advantage and the resource-based view of the firm.
Comparative Advantage is a free market theory pioneered by David Ricardo. The basic idea is that every one could benefit if each country or region made only those things which it could produce relatively cheaper and then trade these goods with other countries or regions for goods which they in turn produce relatively cheaper.
The key idea of the resource-based view of the firm is that firm-specific skills, competences and other tangible and intangible resources are viewed as the bases for the competitive advantage of a firm. The resource-based view has important implications for outsourcing decisions in situations which are concerned with the cost and ownership of resources and economies of scale. Most large Indian IT firms view outsourcing of their own work in terms of resource-based theories, basing their decisions on lack of resources to fill a gap.
However in the case of collaboration amongst firms and networks of firms, it may be more useful to look at outsourcing through the lens of Supply Chain Management and Transaction Cost Economics.
The theories of Transaction Cost Economics originated in the paper “The Nature of the Firm” by Ronald Coase who was awarded the Nobel Prize in 1991. Coase wondered “why a firm emerges at all in a specialized exchange economy”, and “why coordination is the work of the price mechanism in one case and of the entrepreneur in another”. According to Coase, the firm should not be studied in isolation as a production function to be optimized. Instead, the transaction should be made the basic unit of analysis and the firm should be seen as one of several alternative ways of organizing transactions.
Transaction costs consist of search costs, bargaining costs, contracting costs, risk costs, monitoring costs and enforcement costs. The basic premise of transaction cost analysis is that costs associated with transactions vary with the characteristics of the transaction, and different organizational forms are better suited to economize costs for transactions with different characteristics. Transaction cost economics is about finding the most economic organizational form given the characteristics of a transaction.
So viewed through the lens of transaction cost economics, outsourcing should take place when internal transaction costs for producing a good or a service are higher than the market costs, and should be outsourced to a vendor who has a relative advantage in delivering the service.
Fundamentally New Architectures
The rise of the internet, communication technology, collaborative application platforms and international trade co-operation is reducing transaction costs, and is creating an environment and a market conducive for ‘trading in tasks’. The most efficient form of coordination of tasks, especially in delivering IT services, will soon be the work of networks of specialized companies and the market price mechanism rather than of one monolithic company.
John Hagel (of Netgain fame) in his talk on Web 2.0 at NASSCOM 2007, suggested that “Indian IT service companies are natural candidates to define and deploy fundamentally new IT architectures that work from the “outside-in”. In contrast to traditional IT architectures that emerged in the center of the firm and imperfectly extend their reach beyond the boundaries of individual enterprises, we are in desperate need of IT architectures that start with the assumption that the task is to coordinate activities across hundreds, if not thousands of firms. By starting with this perspective, we would need to re-think the nature of transactions and define roles and governance processes accordingly. In fact, we would likely move from today’s transactional architectures to much more helpful relational architectures designed to support enduring and deepening relationships across individuals and institutions”
In March of 2000, Rob McEwen, chairman and CEO of Goldcorp Inc., a precious-metals mining company, unveiled what he called the Goldcorp Challenge. He took proprietary geological data on the company’s Red Lake mine —information normally kept in high-security computer systems—and published it on the Web, inviting geologists, chemists, and computer experts to analyze it and pinpoint likely deposits of gold. The Goldcorp Challenge offered $575,000 as total prize money, with a top award of $ 105,000. In the very conservative and private mining industry, confidentiality and secrecy about reserves and exploration were the norm and Goldcorp’s geologists were appalled at the idea of exposing their super-secret data to the world. McEwen however was majority owner, chairman and CEO, and the challenge went ahead as planned. The top winner was a collaborative effort by two groups in Australia: Fractal Graphics, in West Perth, and Taylor Wall & Associates, in Queensland, which together had developed a powerful 3-D graphical depiction of the mine. In 1996, the Red Lake mine was producing at an annual rate of 53,000 ounces at a cost of $360 an ounce. By 2001, the mine was producing 504,000 ounces at a cost of $59 an ounce. The Goldcorp challenge ultimately yielded 127 locations for drilling and led to a whopping 1.3 million ounces of gold.
Talent Communities
The Goldcorp story is a very good example of how companies can get dramatically better at what they do, and improve faster than their competitors, by working with outsiders whose specialized capabilities complement their own. Different enterprises bring different perspectives and competencies. When these enterprises tackle a problem together, they dramatically increase the chances for innovative solutions.
Illinois Tool Works, a $12.8 billion Fortune 500 company, is a diversified manufacturer of highly engineered components and industrial systems and consumables. It consists of approximately 700 separate companies in 48 countries and employs some 50,000 people. The company started in 1912 as a gear shop and now the group makes everything from nail guns, welders, paint-sprayers to the decorative films that make Victoria’s Secret lingerie shimmer. ITW grew on the back of acquisitions. In 1995 alone, ITW spent well over $6 billion to buy more than 200 companies. ITW routinely breaks up each acquisition into even smaller stand-alone units and gives even business managers nearly complete autonomy, encouraging them to act like CEOs and entrepreneurs.
On average, when ITW picks up a new company, it joins the fold with an operating margin of 9%. Within five years, that average margin jumps to 19%, besting most of its industry peers. ITW’s secret lies in decentralizing authority, and using long tail market segmentation strategies or what ITW calls the 80/20 process at each unit. What this means is focusing exclusively on the 20% of customers or products that provide 80% of sales, and methodically letting go of smaller buyers and lesser product lines. The ITW model seems counter-intuitive. They have over 700 separate companies, and it would seem that they would have more overhead and duplication of costs. But in fact a combination of intense decentralization and long tail market strategies have kept ITW’s decentralized costs lower than the centralized costs of 75% of its competitors.
The Goldcorp story may well be an isolated example of a successful alliance between business entities, and ITW may well be one of a few isolated examples of successfully decentralized businesses. However they are important because they tell us that using collaboration to create value is indeed a commercially successful proposition.
The reality is that creation of successful alliances is an extremely difficult business. Coordinating with outsiders takes its toll. Negotiating terms, monitoring performance, and, if needs are not being met, switching from one partner to another require time and money. In strategic alliances, decision makers must deal with substantial uncertainty, coordination failures and tensions between cooperation and competition, which can lead to opportunism and low performance. We need a deep understanding of alliance implementation, and the processes by which strategic alliances achieve cooperative outcomes and superior performance.
But there is indeed the proverbial pot of gold at the end of the rainbow. Organizations that do learn to manage alliances efficiently and to work with business partners successfully will have immense opportunities in becoming leaders in the creation of talent communities, in using these communities to reduce transaction costs as well as production costs and in commercializing the resultant value network.

This page has the following sub pages.

• Employer Branding and Indian IT firms
• Indian IT: Discriminating against Talent?
• The programmer as a commodity