Conclusion

Why has this industry—which is fundamental to making IT work for corporations, governments, and other organizations—remained “under the radar”? Several factors stand out. First, computer services (commonly referred to as IT services in recent years) are primarily for organizational customers. These services, and the companies providing them, are advertised to the broader public (on TV, on radio, in the popular press, or on the World Wide Web) less frequently than products or services designed for individual consumers. To a degree this is blurred by companies that provide cloud business-to-business and business-to-consumer platforms and services, such as Google and Microsoft, but historically the IT services industry has been—and for the most part it remains—an enterprise-focused business-to-business industry. Second, customer organizations—large corporations, small businesses, government departments and agencies, and nonprofits—tend not to highlight the fact that they outsource some or all of their IT. There is no incentive to highlight that, and (particularly for large corporations or the federal government) doing so might convey a lack of expertise, or an inability or unwillingness to maintain hands-on control and protection of customer/client/constituent information. Third, with Ross Perot from the late 1960s into the 1990s the one possible exception (primarily because he ran for president in 1992 and 1996), it hasn’t had leaders as well-known as those in the computer, software products, and semiconductor industries—for example, Thomas Watson, Thomas Watson Jr., Steve Jobs, Bill Gates, Robert Noyce, Gordon Moore, William Norris, Seymour Cray, Andy Grove, and Larry Ellison.³ Fourth, the industry and its companies tend to gain the interest of journalists and the public only when major government-funded IT services projects (usually large-scale systems integration for the defense or intelligence sectors) are partial or complete failures and the press runs stories focused on taxpayers’ being fleeced. Fifth, it is an industry of many players, with no company ever possessing a market share of 10 percent or more. Much of the work is done by smaller enterprises and even by individual independent contractors. Women made major contributions to launching the industry’s independent contractor brokerage side, and women and smaller businesses have relatively seldom been the focuses of historians of business and technology. Sixth, one component of services is maintenance, which is often overlooked—whether maintenance of machines (repairs and troubleshooting) or code through debugging. Ignoring maintenance and maintenance laborers is endemic in all areas of the history of business and technology. Seventh, IT services is a fluid, rapidly evolving industry, and its boundaries are more porous than those of the hardware products industry or the software products industry (though looking at which companies compete against one another offers important clues and makes defining its segments possible). Eighth, though many of its principal players have substantial presences in Silicon Valley, and though HP was founded and remains headquartered in Silicon Valley, IT is a global, geographically dispersed industry with no one dominant innovative region—leading companies’ headquarters are located in many different places, including Armonk (IBM), Palo Alto (HP), Seattle (Microsoft), Dublin (Accenture), Washington (CSC), Paris (Capgemini), Tokyo (Fujitsu), Mumbai (Tata Consultancy), and Bangalore (Wipro). Large data centers are being built rapidly, across the United States and around the world, and access to land and energy grids are fundamental to locational decisions. And finally, much of what IT services creates—custom software code—is ethereal; it lacks materiality. Software products were boxed and packaged and sold on disks for decades; services are more elusive. Developers of software products (Seymour Rubenstein, Bill Gates and Paul Allen, Mitch Kapor, Bob Frankston, Dan Bricklin) have become famous to varying degrees; leaders of custom applications programming efforts, however large, never have become famous.

One custom programmer, Ellen Ullman, achieved considerable name recognition not from her programming acumen, but by writing a book, Close to the Machine, in which she articulated how the labor done by people who work in programming services (especially for smaller organizations) is invisible to outsiders, oscillates between exhilaration and frustration, and reflects both the creativity and fallibility of its human designers and developers. “The project,” Ullman notes, “begins in the programmer’s mind with the beauty of a crystal … . Code seems lovely in its structureness [sic] … . The world is a calm, mathematical place … . Human and machine seemed attuned to a cut-diamond-like state of grace … .” Next she conveys how this is fleeting: “[S]omething happens … . The irregularities of human thinking start to emerge … . There are dark, unspecified areas.” And finally, “a process of frustration”: “[The] programmer goes back to analysts with questions, the analysts go to the users, the users to their managers, the managers back to the analysts, the analysts to the programmers … . Things are not understood … . The whole graceful structure loses coherence … . A state of grace reveals itself to be a jumble. … [The] human mind is messy.”⁴

Before writing Close to the Machine, Ullman worked—sometimes alone, sometimes on a small team—mainly on programming efforts for nonprofit organizations. For larger projects, the challenges can be even greater. In general, larger programming services and systems integration projects are extremely difficult; that is why a substantial portion of them come in late, over budget, and requiring many hours of debugging, and some are abandoned. This is particularly true for new types of applications in challenging areas such as simulation, health care, advanced real-time logistics, and other fields. From SAGE and SABRE in the late 1950s and the 1960s to a six-year $1 billion plus abandoned (in 2012) Air Force Logistics real-time system modernization effort in the 2000s and the 2010s, software (especially on a grand scale) has proved time and again to be quite hard.⁵

Computer services differ from hardware and software products in blurring the line between producer and users. In consulting and programming services, there is a partnership seeking IT solutions. This is true to a more limited extent with facilities management. With data processing and service bureaus, the services firm often is the end user of the technology even though it is not the end user of the resulting information or data. With many types of computer services, interaction between providers of services and user organizations is frequent and ongoing. Thus, this book contributes to our understanding of computer users and uses, but in a far different way (a different vantage point) than earlier surveys (Cortada’s Digital Hand trilogy) and monographs (JoAnne Yates’ Structuring the Information Age) that are focused on user industries, firms, and other organizations. This book also has some similarities to, but more differences from, FastLane: Managing Science in the Internet World, a book that Thomas Misa and I wrote on the history of the National Science Foundation’s FastLane system. In that book, we closely examined a client (NSF) of a programming services provider (Compuware), but we concentrated heavily on external end users of FastLane (the scientific research community proposing to or reviewing for NSF) and the legacy internal users (NSF staff) of the overall FastLane platform. Compuware is not a primary focus of the book after our brief discussion of its role in programming the earliest iterations of FastLane.⁶

Tabulation rooms at corporations and outsourced pre-computer data processing were the norm when machines for scientific computation—ENIAC, UNIVAC, and IBM’s Defense calculator—entered the scene in the mid 1940s and the early 1950s. The UNIVAC, and the even more business-friendly (in lease price, size, and connection to existing equipment in tabulation rooms) IBM 650, had the potential to serve as tools for data processing, but organizations had to learn to use digital computers for data processing, or learn to effectively make outsourcing decisions (choosing what to outsource and what companies to outsource to). Here, computer consultants were critical and represent the true start of the computer services industry.

The continuities of pre-computer and computer era consulting are clear in the cases of IBM, the accounting firm Andersen and Company’s Administrative Services Division, and the mid-1950s startup consultancies Diebold and Associates and Canning, Sisson, and Associates. The aforementioned consultancies were already well acquainted with the pre-computing tabulation machine data processing technology that many organizations had in place, and were capable of advanced analysis of the economics of computer installations. Andersen and Company, Diebold and Associates, and Canning, Sisson, and Associates often recommended that client organizations continue with pre-computing technology for a while. By providing detailed and trusted technical and economic analysis for clients, they survived and thrived. The computer consulting field of the 1950s certainly included charlatans, with minimal knowledge, who promoted computing in an unrestrained way, but their enterprises did not survive. The reputable consultancies constructed and projected expertise, staying at least one step ahead of the client and putting in the hard work and the study that enabled them to achieve true expertise in digital computing and its applications. Much of their knowledge was gleaned from astute examination of many current and potential client organizations and of ideas and practices that were circulating among clients’ industry peers. Andersen’s Administrative Services Division quickly added a programming services business; Diebold expanded more slowly and diversified at a measured pace. Richard Canning’s business (Roger Sisson left after several years) continued to consult into the early 1960s, but Canning didn’t want to grow the firm and become a manager. Instead, he became the publisher of a monthly newsletter, EDP Analyzer.

The early computer services industry saw the emergence of new segments, new types of services, and new delivery mechanisms. Specialist Automatic Payrolls/ADP stands out for remaining highly specialized in data processing and particularly in payroll processing. Eventually it added transaction processing for stock brokerages as a new and important business area, but payroll processing has always been its main business.

Meanwhile, emerging programming services firms such as Computer Usage Corporation (CUC), C-E-I-R, Inc., and Computer Sciences Corporation (CSC) diversified into service bureaus, data processing outsourcing, and systems integration. Overly rapid growth, cash-flow challenges, and lack of focus (entering hardware and other businesses) outside of IT services contributed to CUC’s downfall. Similarly, a lack of focus and liquidity issues became problematic for C-E-I-R, which was acquired at a time of relative weakness by the computer manufacturer Control Data Corporation. Control Data treated computer services as an important profit center (in contrast to IBM) from shortly after its formation in 1957. That same year, IBM had to split off its service bureau as a subsidiary corporation as a result of a 1956 consent decree, and through the 1960s it bundled its services (offered as “free,” embedded within hardware contracts) as it focused on maximizing revenue and earnings from its hardware business. Owing to concerns about pending Department of Justice antitrust litigation, and recognizing the existence and rapid growth of the computer services industry and the emerging software products industry, IBM executives decided to unbundle many of the company’s software products and some of its services in late 1968. (The unbundling was announced in 1969 and was partially implemented in the early 1970s.)

Selling Service Bureau Corporation to Control Data in a portion of the settlement of CDC’s lawsuit against IBM certainly did not mean that IBM was exiting from services, but it was both a real and symbolic move de-emphasizing services as a recognized business. Yet IBM continued to provide customer engineers, systems engineers, and other services professionals in ever-greater numbers to help customers with IBM equipment and to develop new applications for IBM systems (in order to sell more hardware). Computer services was, without question, a growing activity for IBM, even when it was not a core business for the firm; this has been overlooked in the voluminous secondary historical literature on IBM. And IBM’s Federal Systems Division also provided programming and systems integration services, but much of these services were classified and less visible. IBM’s leaders undoubtedly had less concern that Department of Justice regulators would be scrutinizing services it provided to NASA, the Department of Defense, the Department of Energy, the National Security Agency, and the Central Intelligence Agency.

On the hardware side, at the start of the 1960s, IBM launched its Components Group, soon to be renamed the Components Division. That division manufactured memory and logic semiconductor components for IBM’s systems.⁷ With its computer business and its business and support activities in software, services, and components, IBM was (and remains) a prototypical “Chandlerian” firm—a company with a continuing propensity toward greater vertical integration and toward controlling key (upstream) inputs (materials and components) and (downstream) marketing of products.⁸ In essence, Thomas J. Watson Jr. and his immediate successors strongly favored managerial hierarchies over markets when allocating resources. In the 1960s and the 1970s—like IBM, though to a lesser degree—the entire mainframe industry (including Control Data, Burroughs, GE, Honeywell, and Sperry-Univac) tended toward vertical integration. Though Sperry-Univac outsourced semiconductors, it had a services business, as did CDC, GE, Honeywell, and Burroughs.⁹

In showing contrast to IBM and the mainframe industry’s focus on managerial hierarchies, Dell Computer (later Dell Inc.), as some business historians have highlighted, contributed strongly to and exemplified the use of advanced networking technology for information and logistics to achieve a market-oriented “virtual integration” in the 1980s.¹⁰ The vertical integration of IBM in the 1960s and 1970s and the virtual integration of Dell and lack of integration at its clone manufacturing peers with personal computers in the 1980s and 1990s has become the default time frame for when the computer industry shifts from favoring managerial hierarchies (an earlier dominant period of mainframe manufactures) to a new, post-Chandlerian, market-oriented or network model for personal computers. For years Dell was more innovative with mass customization, supply chain efficiency, and direct marketing than its peers, but it is often overlooked that the whole PC-clone industry used markets or networks for inputs for a product that quickly became a commodity. In this environment, branded semiconductor and software components—“Intel Inside” and “Microsoft Windows”—defined personal computers as much as the assembler’s own branding (Dell, Compaq, Packard Bell, Gateway, Toshiba, Hitachi).¹¹

The timing of this perceived shift from hierarchies to markets in computer “manufacturing” coincides with shifts to market models by the American auto industry and other major industries in the 1980s. The 1980s brought a heightened interest in “lean” production and in the Toyota Production System, exemplified by the joint venture by General Motor and Toyota to partner in building certain GM-branded and Toyota-branded cars in California (which continued until the government-funded post-bankruptcy reformation of GM in 2010).¹² These changes often have been highlighted as the exemplification of an accelerating shift from a manufacturing economy to a service-based economy in the United States and (to a lesser degree) in other developed countries. The IT industry—from hierarchies of mainframes (in the 1960 and the 1970s) to markets for PCs (in the 1980s and beyond)—conveniently fits into such meta-narratives, but only if the focus is on IT’s most conspicuous segment: computers. In computer services—what became IT’s largest industry—changes were underway much earlier.

The computer services industry has always been a mechanism and a force for using markets rather than hierarchies for IT resources. In fact, pre-computer service bureaus facilitated the outsourcing of data processing between the 1930s and the 1950s, before the advent of computer-based business data processing. In the computer era, ADP provided outsourcing of payroll. IBM’s Service Bureau Corporation and other service bureaus offered outsourcing of this type as well as many other business data processing applications. Programming services and systems integration services supplied corporations, government, and other organizations with outsourced custom applications in the 1960s and beyond as IT became more and more important to engineering, design, manufacturing, human resources, logistics, simulation, and many other fields.

Emerging time-sharing firms of the mid 1960s, among them GEIS and Tymshare, allowed users to either supplement or fully outsource their computer processing. Meanwhile, Ross Perot and Electronic Data Systems (EDS) pioneered facilities management as a partial corrective to earlier misallocations of resources for expensive computer systems (internal data processing departments) at customer sites. In essence, EDS and later providers of facilities management represented market mechanisms for more efficient use of computer time among a cluster of computer services’ customer organizations by balancing loads. While IBM was heavily integrated, and its mainframe competitors meaningfully integrated, these companies, and some independent firms, were supplying the computer services or outsourcing options to corporations, the federal government, and other organizations. To be sure, many large companies had substantial computing facilities, some of which did not turn things over to a facilities management services specialist. But even here it is important to recognize the substantial role of external (to the firm) IT labor and expertise. IBM’s share of the computer industry was well over 60 percent. Although many large corporations took possession of IBM systems (generally through leasing), this model was actually a hybrid of hierarchies and markets: IBM’s onsite customer engineers and system engineers provided maintenance, systems integration, analysis, programming services, and debugging.

By the mid 1980s, IBM’s staff of highly trained Systems Engineers numbered more than 20,000 and was focused on making systems useful for IBM customers. In short, while many large companies were buying or leasing hardware and maintaining IT staffs, these firms certainly were outsourcing IT expertise and work, at least in part, to IBM—and, to a lesser extent, to other computer manufacturers, including Control Data, Burroughs, and Honeywell, that provided either bundled or contracted services.

The IT services industry was a facilitator and an exemplar of the great diversity of coordination mechanisms in the US economy (in general, providing strong market options) throughout the entire second half of the twentieth century and beyond—even in the early postwar decades, an era generally associated with the growing managerial hierarchies of major corporations. Ironically, Dell Computer—the company that historians, management scholars, and economists have highlighted as an assembler relying on the market for all the major components of its computers—now is moving partially toward a more integrated model (at least regarding a newfound focus on cloud services) with its $60 billion purchase of the leading data storage company EMC. This purchase (completed in 2016) was the largest IT acquisition in history.¹³ Reminding us of the IT industry’s diversity of coordination, rather than a trend toward greater integration, HP’s management recently headed in the opposite direction, dividing the longtime IT giant into two separate corporations—HP Enterprise (services) and HP Inc. (computers and printers). HP Enterprise has sought to increase its scale and customer base in merging with the remaining commercial computer services business of CSC. And IBM, while still heavily integrated, is partnering with Apple and other former competitors and is concentrating more and more on cloud services and data analytics. Twenty years from now, it probably will be a very a different company—it may have retired completely from manufacturing hardware, and its organizational structure may look more like the split-off HP Enterprise or the consulting-focused Accenture than like today’s IBM, with its divisions for services, hardware, and software.

Thanks to low barriers to entry, IT services has long been a low-margin, fast-growth business. Parts of the industry’s contributions that long ago were considered “cutting edge” (for example, business data processing for accounting, payroll, and human resources) have become commonplace and less profitable. Areas ripe for innovation are industry-specific verticals where new software and customized systems are created for the financial, petroleum, health care, aerospace/defense, biotechnology, and many other industries.

Much of the quite small management-and-economic theory-focused literature on innovation in IT services is presented more broadly under the rubric of “services innovation.” In an influential article published in 1986, Richard Barras posited that services might have a “reverse product cycle” characterized by efficiency gains, quality gains, and new services.¹⁴ That idea could be useful in regard to the business data processing previously done in tabulation rooms and later accomplished by computers, but it seems far less useful for understanding major systems integration efforts such as SAGE and SABRE. IT not only provides opportunities for efficiency gains and quality control; it also creates new possibilities. Increasingly, many of those possibilities appear to be at the intersection of mobile computing, cloud services, and the application of artificial intelligence in the dynamic field of data analytics. Here innovations are not merely for process improvements but for engineering new knowledge to help firms and organizations.

Further, cloud platforms are blurring the gap between what is a software product and what is a service. Is a base of standard code always a product if interaction with the supplying IT company’s staff inevitably is critical to its implementation with a particular organization and its unique setting and circumstances? In one sense it is assistance with a product; in another sense it is creating something unique from a base where all is intertwined in a system of interaction and modification by a services provider working closely with a client. Applications that are available on the go—such as Salesforce.com’s cloud CRM platform and services—bridge mobile, cloud, and software platforms and customizable applications to offer new possibilities. This has made Salesforce.com one of the fastest-growing IT startups in recent decades and by far the largest firm specifically founded to provide cloud computing enterprise services. Cloud services, in general, bring together data as never before, continually creating new possibilities for big data analytics, a differentiation area already yielding high margins and likely to continue to do so.

The number of IT services employees in the US may soon reach a plateau, and some of the largest players have decreased their domestic headcounts as they have been expanding their workforces based in India, the Philippines, and other lower-labor-cost countries. IT services, however, is a diverse industry, and many smaller players in IT services have grown domestically. Except during the Great Recession (2008–2009), domestic employment in this industry has gradually increased, and a major portion of higher-margin work in cloud computing and in analytics for US firms is done by IT services companies in the US—the host nation of many data centers and server farms. This bodes well for talented and location-flexible US-based IT professionals. To what extent US multinationals will look to lower-cost labor in India and in other countries for IT services for cloud businesses and infrastructure remains to be seen. (Amazon, Microsoft, and other once overwhelmingly domestic employers are now increasing their operations in India.) It also remains to be seen whether disturbing trends with the declining levels of women’s participation in IT over the past three decades (both in education and industry) can be reversed.

Some of the major players in the IT services industry—including IBM, HP, and Accenture—have formal efforts/programs/goals to boost the workforce percentage of women workers/professionals/managers. To facilitate positive outcomes, there likely will need to be changes within corporations, government, colleges, and universities (where there is a far lesser percentage of women majors in computer sciences after a late 1980s peak), K–12 education, professional organizations, trade associations, and other institutions, with regard to gender and underlying cultures of computing (and more broadly, technology). This book has shown that women had some opportunities and made major contributions to computer services but also faced some considerable hurdles in reaching upper levels of technical and managerial leadership in the industry (especially at midsize and large firms). Computer Usages’ first four (hired) programmers (in 1955) were all women, yet six years later, as the programming force expanded to dozens and five hierarchical job classes had been formed in programming, the programmer labor force at the firm was predominantly male and exclusively male within the top job classification of “principal analyst.” And five years after that, the one woman who had been on the board, Louise Greene was no longer a board member and the company had an exclusively male board. Before seeking to control their careers as entrepreneurs, Grace Gentry, and others were discriminated against at large organizations. Sometimes this was explicit, other times it took the form of conversations and informal meetings in male gendered spaces. Hopefully in the future there will be deeply researched studies focused on gender and the various IT industries (computer, software products, and computer services) in the US and worldwide, as well as internal IT operations of corporations, government, and other organizations. Marie Hicks’ recently published book Programming Inequality: How Britain Discarded Women Technologists and Lost Its Edge in Computing, provides a rich example in analyzing the history of women IT workers in Great Britain and the consequences of discrimination.

In 2003, Nicholas Carr published in the Harvard Business Review a controversial article which he expanded into a short book the following year.¹⁵ The provocative and somewhat deceptive titles of the article and the book were, respectively, “IT Doesn’t Matter” and Does IT Matter? In the book, Carr looked at the approximately $2 trillion spent annually on IT (including both the major industries of IT—computers, software products, computer services, networking/communications—and in-house IT personnel) and argued that IT’s strategic importance for corporations had dissipated. Carr’s argument—a meaningful departure from the general question “does IT matter?”—really focuses on whether investing in IT for competitive advantage makes sense for most companies in view of the competitive landscapes of their industries.¹⁶ In Does IT Matter? and in a subsequent book, The Big Switch: Rewiring the World, From Edison to Google, Carr characterizes cloud services as a massive commodification of IT to the point where it essentially is a utility—like electricity.¹⁷ There is certainly a continuing trend toward cloud services; however, as is true of computer services overall, the cloud space is extremely complex and varied, and it has quite strong elements of both commodification and differentiation. Ever since Automatic Payrolls/ADP’s largely commodity payroll processing and SDC’s highly differentiated programming and systems integration work on SAGE in the second half of the 1950s, in the early 1960, and in subsequent years, the computer services industry has always been a mixture, and it probably will remain so for a long time.

Carr cites a 1997 survey of CEOs and directors of large American and European corporations in which respondents indicated that they believed that, by the year 2000, 60 percent of their IT initiatives would be for “achieving competitive edge,” not just “catching up or staying afloat.”¹⁸ Realistically, how many leaders of very large corporations that are spending hundreds of millions or billions of dollars annually on IT are going to admit to themselves or others that their firm is just trying to tread water, or remain in the middle of the pack, with IT (whether internal, outsourced, or a hybrid) or any other high-expenditure function? When the survey Carr cites was conducted, the dot-com bubble was expanding rapidly. Attitudes were temporarily at a fever pitch; they probably would have been more subdued after the dot-com bust of the early 2000s. In 2004, when Carr’s book was published, the (information) technology-laden NASDAQ market was still down 60 percent from its March 2000 high. In areas of considerable investment in technology and labor, company leaders want their firm to achieve excellence and to have a competitive edge (if only a small one), even when not engaging in outsized IT investment relative to competitors. Keeping up with IT in industries such as finance, petroleum, automobiles, banking, insurance, and retail will always be very expensive, even when a corporation is not trying to one-up its rivals with IT. Productivity gains from overall IT spending by corporations and other organizations have long been debated (and isolating one factor’s role with productivity is notoriously difficult), but questions about the productivity paradox that were widely cited in the 1990s (questions as to whether IT investments led to comparable gains in productivity) have seemingly been put to rest by studies showing substantial returns on IT investment.¹⁹

Carr’s analysis also pays relatively little attention to the financial and reputational costs that major breaches of computer security can bring.²⁰ Over the past decade, high-profile security breaches or hacks have hurt Sony Pictures, Home Depot, Target, AshleyMadison.com, eBay, JPMorgan Chase, and many other corporations. Significant breaches also have been suffered by the Department of Defense, the Internal Revenue Service, the Office of Personnel Management, many state and local governments, and nonprofit organizations.

Appropriate information technology infrastructure and investment (for a particular organization in its particular industry or environment), the right technological tools coupled with proper procedures and practices to best ensure security, and taking part in an ecosystem characterized by knowledge-circulating conduits for information technology all are extremely important. The expertise and knowledge-circulating functions of providers of IT services—giants such as IBM, HP, Accenture, Fujitsu, and Tata Consultancy, smaller players, and specialists in security products and services such as Palo Alto Networks—can be critical to organizations and may require major IT investments. In short, IT, understood within its broader systems of organizational structures and users, most certainly does matter.

Carr cites Max Hopper—American Airlines’ longtime top IT manager, who directed the evolution of SABRE in the 1970s—as “bravely” and insightfully seeing the writing on the wall in 1990 by predicting that IT would come to be seen as “more like electricity or the telephone network than as a decisive source of organizational advantage … . In this world, a company trumpeting the appointment of a new chief information officer will seem as anachronistic as a company today naming a new vice president for water and gas … .” Max Hopper (who I had the privilege of meeting and chatting with in 2004)—unlike Nicholas Carr or myself—was old enough to remember, and probably was influenced by, the extensive writings and rhetoric about a future computing utility in the late 1960s, when time sharing was still a young technology and was sometimes spoken of in idyllic, far-reaching, and utopian terms. Several decades later, Hopper felt these utopian dreams were coming true.

Chief Information Officers, however, are more important than ever, regardless of where the boundaries of the firm are drawn with IT—outsourcing wisely, like internal IT operations, takes great knowledge and skill.²¹ And in addition to CIOs, most major corporations now have Chief Security Officers. Both CIOs and CSOs will become increasingly important to corporations and other organizations in the years and decades ahead.

The trend toward outsourcing of IT continues to gain momentum, and for decades the IT services industry has grown far faster than the overall economy. Organizations will purchase a range of standardized information technology platforms and applications, as well as custom solutions. Cloud computing unquestionably includes some characteristics of the old computer utility concept, particularly infrastructure as a service, but many customized cloud services will be purchased by organizations, many to their advantage, but undoubtedly some purchases will be misallocations of resources. Cloud analytics, which probably will continue to grow rapidly, offers new possibilities for making the most of ever-larger stores of data. Computer security will remain a challenge, and IT services companies will continue to invest heavily in tools and techniques to protect themselves and their clients. Computer services, as has been true since the start of the industry more than six decades ago, will be characterized by some homogeneity (commodification), much heterogeneity (differentiation), a range of successes and failures, many uncertainties, and very rapid change.

Notes

2. The important work of Martin Campbell-Kelly (both his single-authored scholarship and his co-authored scholarship with Daniel Garcia-Swartz) stands as an exception—two chapters in From Airline Reservations to Sonic the Hedgehog, Campbell-Kelly’s book on the software products industry, focus entirely or largely on services, and Campbell-Kelly and Garcia-Swartz’s excellent, concise, and engagingly written history of the entire IT industry, From Mainframes to Smartphones, is insightful. Because the survey by Campbell-Kelly and Garcia-Swartz covers the six-decade history of the entire global computer industry—computers/hardware, software, and services—it gives more attention to the industry’s hardware side. It is also more an economic history than business history. In contrast, this book is more focused on business history, and offers case studies of numerous IT firms and their evolving strategies—including some in industry sectors largely ignored by other scholars (such as consulting, facilities management, data processing services, and independent contractor brokerages).

20. Carr (2004, 2008). In the former, little mention is made of security. In the latter, security makes several bit appearances in discussions of malicious bots and hacks, but it is not well integrated into Carr’s argument as to what this might mean for the commodity cloud or for the IT as a utility era he foresees. In general, a degree of customization and diversity makes things harder for malicious attackers. And corporate and government spending on technology, but also on education, behavior, and processes, will have to increase dramatically (as it has already begun to do). I co-led a major history of computer security project sponsored by the National Science Foundation (2012–2015), and am currently writing a book on the history of computer security under contract for the MIT Press.