SOFTWARE INDUSTRY, consists of that part of computer programming activity that is traded between software-producing organizations and corporate or individual software consumers. Traded software represents only a fraction of domestic software activity, whose extent cannot be reliably estimated, since much computer programming takes place within firms and its value is not captured by the industrial census or software industry analysts. According to the industry analyst INPUT, in 2000 the U.S. market for traded software was $138 billion (Table 1). The U.S. software industry is a major exporter, and the total revenues of the top 500 U.S. software firms in the year 2000 were $259 billion, according to the trade publication Software Magazine.
The traded software industry consists of three main sectors: programming services, enterprise software products, and shrink-wrapped software products. These three sectors became established in the mid-1950s, the mid-1960s, and the late 1970s, respectively, in response to the technological opportunities and the business environment of the time. The most successful firms developed specialized
|U.S. Software Market (User Expenditures in $ millions), |
|source: Courtesy of INPUT.|
capabilities that enabled them to prosper within their sector; however, this specialization made it difficult to move into other sectors, and very few firms have been successful in more than one software sector. It should be noted that the software industry is not confined to independent software vendors, but also includes computer manufacturers such as IBM, Unisys, and NCR who supply programming services and software products alongside their hardware offerings and are among the largest software suppliers. These are sometimes referred to as "captive" markets because computer users have relatively little choice in the supplier of basic operating software for corporate systems.
The United States has been the world leader in the software industry throughout its history, and today accounts for half of global revenues overall, and an estimated three-quarters of the software products market. A notable feature of the industry is its low concentration: there are many thousands of software firms in the United States and throughout the world, but relatively few—mostly American—global players.
The first commercial electronic computers—usually known as "mainframes"—were sold in the early 1950s. They were very expensive, typically renting for $100,000 a year. Most computer-owning corporations undertook their own program development and operations, for which they maintained a staff of up to thirty or forty individuals. This was not a disproportionate expense in relation to the overall costs of running a computer.
By the mid-1950s, however, mainframe prices had fallen significantly, and computer use diffused rapidly—the national computer stock rising from 240 mainframes in 1955 to over four thousand by 1960. Owners of these more moderately priced computers were often unwilling to recruit a permanent programming staff, preferring instead to commission programs from software contractors. Many of the early programming services firms were established by programming entrepreneurs to satisfy this demand. The first such firm, the Computer Usage Corporation (CUC), was incorporated in New York in 1955 by two former IBM programming employees, and the firm initially specialized in developing technical applications for the oil and engineering industries. The capital barriers to software contracting were (and remain) very low, and it was often said that all one needed was "a coding pad and a pencil." The most important capability was the technical knowledge of the principals, usually acquired through working with a computer user or manufacturer. Several dozen firms entered the programming services industry in the second half of the 1950s. In a majority of cases, the firms specialized in particular technical applications, or within a vertical market such as financial services, retail, or manufacturing.
A very different type of entrant came into programming services in the mid-1950s, specializing in the construction of very large programs that were beyond the technical capability of even the largest and most sophisticated users. The first firm of this kind was the Systems Development Corporation (SDC), a subsidiary of the RAND Corporation, Santa Monica. SDC was incorporated in 1956 to develop the programs for the giant SAGE air-defense system. The programs SDC developed for SAGE were unprecedented in size, consisting of more than a million computer instructions. SDC employed several hundred programmers, estimated at the time to be perhaps halfof the nation's programming manpower. SDC also trained hundreds of individuals to become programmers. There was, however, a rapid turnover of staff, as experienced programmers left for more remunerative employment in the private sector. At the time, SDC was hailed as the "university for programmers" and it was said that in the 1960s, SDC alumni were to be found in almost every major software firm in the country.
SAGE was a "real-time" system, in which the computer lay at the heart of an information system that responded instantaneously to external events. As the U.S. government deployed more and more real-time defense systems in the late 1950s and 1960s, systems integrators such as TRW, MITRE, General Electric, Westing house, Hughes Dynamics, and Lockheed began to develop expertise in software construction. Real-time technologies were hugely expensive to innovate but once established by the military, they quickly diffused into the civilian sector in applications such as airline reservations and on-line banking. When Europe and the rest of the world began to catch up in the 1960s, American independent software firms and the programming services operations of computer manufacturers had a strong first-mover advantage.
By the late 1960s, the most successful of the start-up software contractors had become significant firms. For example, by 1967 CUC had 700 employees, offices in twelve U.S. cities, and annual sales of$13 million. CUC, and firms like it, now offered a broad range of computer services that went well beyond program writing. Another firm, the Computer Sciences Corporation of El Segundo, California, established in 1959 by five programmers to write software for computer manufacturers, grew to become one of the largest computer services firms in the world (which it remains, with revenues in 2000 of$9.4 billion, and sixty-eight thousand employees worldwide).
Nonetheless, giant firms are the exception and the programming services industry is remarkably lacking in concentration. By the late 1960s there were several hundred U.S. programming services firms, but less than fifty of them had as many as a hundred employees. Today, there are several thousand software contracting establishments, but their average size is less than a dozen staff. Of these only the smallest percentage will become global players.
Enterprise Software Products
The 1960s saw an explosion in computer usage in the United States and worldwide. Computer technology evolved dramatically—transistors replaced vacuum tubes, and then microelectronics superceded discrete transistors—with consequent improvements in speed, capacity, reliability, and price. In the United States the computer population grew from 4,400 in 1960 to 46,000 by the end of the decade (63,000 worldwide). The concept of packaged software arose as a technological solution—almost an historical necessity—to the problem of supplying software for the expanding computer population, which by the mid-1960s had begun to outstrip the programming manpower needed to write custom software for each individual installation. A software package was a pre-written program designed for a particular industry, or for a common application such as payroll processing or inventory management. At first software packages were supplied at no cost by computer manufacturers, as part of the bundle of services needed to operate a computer, and which included customer training, field engineering, and so on.
In the mid-to late 1960s a number of established programming services firms began to sell software packages. The packages were usually derived from software assets developed in programming-services contracts, and were called "products" to distinguish them from computer manufacturers' software packages, in order to connote a degree of customer support and service that the manufacturers generally failed to provide with their free packages. Two early examples of software products were Autoflow and Mark IV. Autoflow, a program that assisted computer users with software documentation, was introduced in 1965 by Applied Data Research (ADR), which was founded in Princeton, New Jersey, in 1959. Mark IV was an early form of database launched in 1967 by Informatics (founded in Woodland Hills, California, in 1962). Although Autoflow and Mark IV were perhaps the two best-known software products of the 1960s, they had each achieved only a few hundred sales by the end of the decade.
The manufacturers' distribution of "free" software packages was a major restraint on the growth of the software products industry, because it was extremely difficult to compete against free packages by providing added value that justified a purchase price typically offive thousand to fifty thousand dollars. As a result of antitrust pressure, and a private lawsuit from ADR, in 1970 IBM "unbundled" its software and services, converting many of its software packages into paid-for software products. Unbundling gave a major fillip to the industry; ADR and Informatics, for example, tripled their sales in as many years. During the 1970s, several hundred of the existing programming services firms and many new ventures entered the software products industry. New entrants included firms such as Computer Associates and Oracle, later to become leading players in the enterprise software industry. However, the 1970s was a somber decade for the software industry, with capital shortages following the stock market crash of the late 1960s, followed by the computer recession of 1970–1971. Hence growth in the 1970s was modest, and total industry sales did not exceed $1 billion until 1978 (a year in which IBM's revenues were $17 billion, for comparison).
During the 1980s, the software products market finally matured and grew at a sustained annual compound rate of 30 percent—from aggregate sales of$2.7 billion in 1980 to over $30 billion by the end of the decade. As with programming services, the software products industry was low in concentration. For example, a survey in 1982 showed that the top fifty or sixty firms accounted for only 50 percent of sales, leaving the remainder to approximately two thousand medium and small firms. (By contrast, in the mainframe computer industry, less than twenty firms accounted for virtually the entire industry, and one firm, IBM, for more than 50 percent.) The software industry was (and is) sometimes characterized as being like "boulders, pebbles, and sand"; that is, a few tens of global players, a few hundred second-tier firms, and thousands of very small firms with a dozen or fewer employees.
The leading firms have generally grown through consolidation or by dominating a particular software genre through organic growth. Consolidation has been particularly important in gaining market share in the software industry for two reasons: it has proved extremely difficult to imitate a successful software product that may have taken years to evolve, and software products have proved insensitive to price, buyers valuing reliability and security of supply above cost. Among the consolidators, Computer Associates, founded in Islandia, New York, in 1976, has been the most prominent and successful. Following its initial public offering in 1981, Computer Associates set out on a course of acquisitions that made it the largest software vendor by 1990 and it has retained its place as one of the top three vendors ever since (usually jockeying for position with Microsoft and Oracle). Computer Associates' acquisitions have included other consolidators, so that if one were to construct a "family tree" of the firm it would contain several hundred names (including ADR and Informatics, mentioned above).
By the 1990s, two major software genres accounted for perhaps a half of all corporate software sales: relational database software and enterprise resource planning (ERP) software. Relational database technology emerged in the early 1970s in the research environment of IBM's San Jose Research Laboratory and the University of California, Berkeley. Relational technology was a major, though technically challenging, advance over earlier database systems. The technology was first exploited by northern Californian software entrepreneurs, including Oracle, founded in Belmont, California, in 1977. The region still remains the world center of relational technology. Oracle has consolidated its early start advantage, out-maneuvering and out-growing all competitors, frequently vying with Computer Associates as the number-two software company. ERP software emerged in the 1980s as a single-product solution to replace the aggregation of numerous application products that computer users typically had to use in the 1970s and 1980s. The leading vendor is SAP, a German firm, which invented the ERP concept in the early 1980s; although there are now several U.S. competitors, none has yet overcome SAP's first-mover advantage. SAP is the only non-U.S. software-product firm in the top ten (and one of only a handful in the top 100).
The invention in 1971 of the inexpensive microprocessor—a computer on a single microchip—transformed the computer, creating a consumer product from what had previously been a costly capital good. Microprocessors were used in both videogame consoles and personal computers, and a "shrink-wrapped" or "boxed" software industry developed in the 1970s to satisfy the demand for programs for the new computer products. Shrink-wrapped software products were distinguished from enterprise software goods by low prices, high sales volume, and different distribution channels. Consumer software typically sold in tens or hundreds of thousands of units, at a price of a few hundred dollars, at most. Shrink-wrapped software was sold through retail outlets and mail order, with little or no after-sales service, compared with the direct sales forces and account managers employed by enterprise software vendors.
One of the first firms to catch the wave of personal computing was Microsoft, founded in Albuquerque, New Mexico, in 1975 by Bill Gates and Paul Allen. Microsoft specialized in the basic operating software for what were then still known as micro-computers, and which mainly sold to hobbyists. Mass-market personal computers, such as the Apple II and the Tandy TRS-80, arrived in 1977– 1978. Within a couple of years two software applications, the word processor and the spreadsheet, made the personal computer generally useful. The word processor concept had existed in the corporate computer world, and many entrepreneurs simultaneously hit on the idea of creating a product for the personal computer market. However, one product, WordStar (produced by a San Rafael-based start-up, MicroPro International), secured a commanding share of the market. By contrast, the first personal computer spreadsheet, VisiCalc, produced by Software Arts of Cambridge, Massachusetts, had no clear precedent in mainframe computing, and it took the world by storm, initially having the market entirely to itself, although within a couple of years it was imitated by dozens of "clones."
WordStar and VisiCalc were the two top-selling programs, each with sales of more than half a million copies by 1983. WordStar, VisiCalc, and their competitors were known as "productivity applications" because they became the principal tools of corporate information workers. Although productivity applications accounted for around half of software sales, there were many other software genres, each with their leading products—desktop publishing (for example, Adobe Systems' PageMaker), computer-aided design (for example, Autodesk's Auto-CAD), personal finance (for example, Intuit's Quicken), and many more.
The tendency for a single product to dominate a software genre, despite many worthy imitators, is a notable feature of the personal computer software industry and is mainly attributed to "network effects." Because a community or network of users values the ability to share and exchange files and documents, new users tend to adopt the software with the biggest community of users—this favors the most popular product to the detriment of its competitors. The pecking order of top software products tends to change only when there is a major discontinuity in the personal computing environment, which temporarily overcomes these network effects. This first happened following the introduction of the IBM-compatible PC in August 1981. The IBM PC was more powerful than previous desktop computers, and IBM's imprimatur legitimated its use in corporations, and it soon accounted for 80 percent of new sales. Two new productivity applications eclipsed VisiCalc and WordStar, and dominated the platform for a decade: the 1-2-3 spreadsheet by the Lotus Development Corporation (Cambridge, Massachusetts) and the WordPerfect word processor by the Word-Perfect Corporation (Orem, Utah). Microsoft, by now relocated to Gates and Allen's home town of Seattle, Washington, developed the operating software for the IBM computer, earning enormous revenues that fuelled its growth.
A seismic shift occurred in the personal computer landscape with the introduction of the Microsoft Windows 3.0 operating system in 1990. Because the productivity application incumbents, Lotus and WordPerfect, did not have Windows-compatible products at this time, Microsoft was able to push its own packages, Excel and Word, with very little competition. At the same time, the company began to bundle its productivity applications into a single product called Microsoft Office. Within a couple of years, Microsoft dominated the markets for both personal computer operating software and productivity applications. This precipitated a restructuring of the industry as competitors repositioned and merged in order to offer comparable "office suites." By the mid-1990s, office suites and individual productivity applications sold some fifty million units annually, with perhaps three-quarters being Microsoft products.
Microsoft's ubiquity and monopolistic practices brought it to the attention of the antitrust division of the Department of Justice in 1995. This exclusive focus on Microsoft has been surprising, since there are comparable monopolies in the enterprise software sector. Moreover, while Microsoft certainly dominates most of the software genres in which it competes, the firm accounts for no more than 10 percent of total software industry sales.
The Internet Era
The diffusion of the PC in the 1980s dramatically changed the working lives of office employees: senior managers began to type their own memoranda, while junior executives spent a disproportionate amount of their days tinkering with spreadsheets. The typewriter was consigned to the dustbin of history, while typists were promoted to being general administrators.
Since the mid-1990s, the Internet has had an equally dramatic effect on the lives of office workers and increasingly the domestic computer user. Although the Internet had been in widespread use in technical communities since the early 1980s, it was the introduction of the World Wide Web in the early 1990s that made the Internet accessible to ordinary users. The enabling software technology of the World Wide Web was the Web browser used in a desktop computer, of which most users are conscious, and the invisible software in "servers" that made Web pages available to remote users on demand. The first and most important supplier of software for the Web was the Netscape Communications Corporation of Mountain View, California, founded in 1994. Run by a 24-year-old wunderkind, Netscape grew from nothing to a billion-dollar corporation in two years. In 1996, however, Microsoft introduced a competing product, Internet Explorer, which quickly achieved a dominant market share. Microsoft's hardball tactics in this achievement added weight to its antitrust prosecution.
At first, the Web was a largely passive experience, but in the late 1990s it became increasingly interactive and participative, with the provision of financial services, travel information, entertainment including pornography, auctions and retail services, and information products of every kind. In a couple of years new brands became household names—Motley Fool, Ameritrade, Yahoo, Travelocity, Amazon, and many more. Some of these are new enterprises, while others are new initiatives from old established firms. Just as the Web is changing the old order of information services and retailing, it is even more profoundly changing the world of software. For example, the metaphor of "shrink-wrapped" software is breaking down as software products are increasingly supplied by electronic download. Some industry observers predict that the concept of a software product or artifact will become obsolete, and programs will be supplied online, on demand, and metered according to use. Whether or not this comes to pass, the phrase software industry will continue to be the collective term for firms engaged in supplying programming goods and services, in whatever way the technology of the day demands.
Baum, Claude. The System Builders: The Story of SDC. Santa Monica, Calif.: Systems Development Corporation, 1981.
Campbell-Kelly, Martin. A History of the Software Industry: From Airline Reservations to Sonic the Hedgehog. Cambridge, Mass.: MIT Press, 2003.
Cusumano, Michael A., and Richard W. Selby. Microsoft Secrets: How the World's Most Powerful Software Company Creates Technology, Shapes Markets, and Manages People. New York: Free Press, 1995.
Freiberger, Paul, and Michael Swaine. Fire in the Valley: The Making of the Personal Computer. 2d ed. New York: McGraw-Hill, 2000.
Herman, Leonard. Phoenix: The Fall and Rise of Home Videogames. Union, N.J.: Rolenta Press, 1994.
Hoch, Detlev J., Cyriac R. Roeding, Gert Purkert, and Sandro K. Lindner. Secrets of Software Success: Management Insights from 100 Software Firms Around the World. Boston: Harvard Business School Press, 2000.
Liebowitz, Stan J. and Stephen E. Margolis. Winners, Losers and Microsoft: Competition and Antitrust in High Technology. Oakland, Calif.: The Independent Institute, 1999.
Manes, Stephen, and Paul Andrews. Gates: How Microsoft's Mogul Reinvented an Industry and Made Himself the Richest Man in America. New York: Simon and Schuster, 1994.
Meissner, Gerd. SAP: Inside the Secret Software Power. New York: McGraw-Hill, 2000.
Mowery, David C., ed. The International Computer Software Industry: A Comparative Study of Evolution and Structure. New York: Oxford University Press, 1996.
Organization for Economic Cooperation and Development. Software: An Emerging Industry. Paris: OECD, 1985.
Petersen, W. E. Pete. Almost Perfect: How a Bunch of Regular Guys Built WordPerfect Corporation. Rocklin, Calif:. Prima, 1994.
U.S. Department of Commerce. A Competitive Assessment of the United States Software Industry. Washington, D.C.: U.S. Department of Commerce, 1984.
Wilson, Mike. The Difference Between God and Larry Ellison: Inside Oracle Corporation. New York: Morrow, 1997.