Figure 1.1 Frances Bilas (right) and Betty Jennings (left) adjusting program settings on the ENIAC master programmer in 1946. Courtesy of Charles Babbage Institute, University of Minnesota.
On February 14, 1946, nearly six months after the end of World War II, the US Department of War officially dedicated the Electronic Numerical Integrator and Computer (ENIAC) at the University of Pennsylvania. The machine, a 30-by-60-foot behemoth containing about 18,000 vacuum tubes, had been operating for three months. Various types of analog and electromechanical digital computers had preceded it, from Vannevar Bush’s Differential Analyzer and George Stibitz’s relay-based Complex Number Computer in the 1930s to Howard Aiken’s Mark I and Konrad Zuse’s Z3 in the early 1940s. But the ENIAC was the first meaningful electronic digital computer.1
The US Army provided more than $400,000 for the design and development of the ENIAC, a project led by the physicist John Mauchly and the electrical engineer J. Presper Eckert at the University of Pennsylvania’s Moore School of Electrical Engineering. The Department of War authorized the funds to create a machine of unprecedented speed and accuracy for calculating ballistic firing tables. From the launch of the project through the end of the war, the ENIAC effort was little known, having been classified as “confidential.”2 The New York Times reported that at the dedication officials of the Department of War characterized the ENIAC as a “tool with which to begin to rebuild scientific affairs on new foundations.”3 In December 1945, unbeknownst to spectators, the ENIAC had been used to calculate equations on thermonuclear reactions in support of research for developing a hydrogen bomb.4
Scientists, engineers, government officials, and the press all understood the ENIAC and other electronic digital computers of the second half of the 1940s to be advanced scientific calculating machines. Common metaphors for the ENIAC and other digital computers of the time were “giant brain” and “electronic brain.”5
In 1946, Eckert and Mauchly left the Moore School and founded the Electronic Control Company (soon renamed Eckert-Mauchly Computer Corporation) to design and manufacture digital computers. That same year, in St. Paul, former US Navy cryptographers and investors launched Engineering Research Associates to engage in naval electronics contracting and to design and build digital computers. In their first years, these pioneering computer companies focused solely on building computational machines and other electrical equipment for the scientific, engineering, and defense markets.
Figure 1.1 Frances Bilas (right) and Betty Jennings (left) adjusting program settings on the ENIAC master programmer in 1946. Courtesy of Charles Babbage Institute, University of Minnesota.
At a session of the 1949 annual meeting of the American Institute of Electrical Engineers (AIEE) that lasted nearly four hours, engineering professionals discussed the future possibilities for digital computers. Among the list of potential applications, the engineers postulated that digital computers could be useful for flying airplanes, tuning televisions, forecasting the weather, and automating manufacturing. This distinguished group also debated what was required for a computer to “have a will of its own” and what constituted a “thinking” device.6 At the meeting, Jay Forrester, an engineer from MIT’s Servomechanisms Laboratory who was orchestrating the expansion of an analog flight simulator project into a successful effort to build the first real-time digital computer, declared “present efforts in digital computing are only exploratory, the engineering effort which will ultimately be justified may exceed that devoted to the wartime development of radar.”7 While Forrester and others understood the limitations of comprehending a new technology in its infancy, this group of distinguished AIEE engineers nevertheless correctly identified important future applications, recognized the importance of computers to questions being raised in the young field of cybernetics, and anticipated issues that would be central to the future field of artificial intelligence. Absent from these engineers’ reported discussions was the possibility of applying digital computers to data processing for businesses.8
In the second half of the 1940s and in the early 1950s, the state of the art in business data processing was the use of electromechanical punch card tabulating machines from the office machines giants International Business Machines (IBM) and Remington Rand. IBM began under the name Computer-Tabulating-Recording Company (C-T-R) in 1911. It was a leading producer of punch card tabulation machines from its origin well into the 1960s (when selling and leasing computers became its primary business).
The transition from punch card tabulation to computers took time and was uneven among various data processing machine user industries. Before 1960, Remington Rand was IBM’s primary competitor in the market for punch card tabulation machines. As historian Thomas Haigh has pointed out, in the early 1950s it was far from clear that the electronic digital computer would be the key electronic product for business data processing.9
Like IBM’s leaders, top executives at IBM’s competitors (Remington Rand, Burroughs, and National Cash Register) fully understood by the late 1940s that the future of office machines lay in electronics.10 Those firms, however, lacked the extensive financial resources that IBM was able to devote to greatly expand its electronics research and development. Recognizing the market for scientific computing and hopeful of applications of digital computers in business, Remington Rand acquired the Eckert-Mauchly Computer Corporation (in 1950) and Engineering Research Associates (in 1952) in stock-based deals.11
While Remington Rand’s Universal Automatic Computer, (UNIVAC)—first made available in 1952—was designed and marketed as a machine for both science and business, in most configurations the price exceeded $1 million, an amount few corporations could justify spending for new data processing technology. The much smaller IBM 650 sold for between $200,000 and $400,000, and leased for as little as $3,500 a month.12 IBM had more than 2,000 IBM 650 installations (sales and leases) during the 650’s eight-year production run, compared with fewer than 50 installations for Remington Rand’s UNIVAC. IBM also continued to generate substantial revenue from selling and leasing its pre-computer punch card tabulation systems throughout the second half of the 1950s and into the 1960s.13
In the 1950s, a decision to acquire a digital computer (or multiple computers) appeared daunting for most corporations and other organizations. Mainframe computers were extremely expensive and were technological and institutionally disruptive. Moreover, the economic payoff digital computers might provide for business data processing was uncertain. Pre-computer punch card tabulation equipment, which most major American corporations had acquired between the late 1920s and the early 1940s, also had been costly, but significantly less costly than digital computers. Many corporate decision makers likely had some knowledge of the challenges of getting punch card tabulation departments up and running. By the advent of the first viable commercial computers on the market (in 1952), established routines were in place within corporate data processing departments—from the technology itself and its spatial location to the personnel, their knowledge and skills, and organizational culture. The thought of completely uprooting these organizational structures and routines in favor of new technology tended to be overwhelming—and represented just the start of the many challenging questions to answer: What would be the first applications? What formal training procedures would be launched? What new positions would be created? What would be done in house? How would corporate or organizational reporting responsibilities change?
In opposition to the many reasons for corporate leaders’ apprehension with taking the leap to an expensive and unproven new technology for business data processing was the fear of being left behind. By the mid 1950s, computer manufacturers’ sales staffs, journalists, consultants, and others had fueled this fear by serving as evangelists for the inevitable electronics-based (and sometimes specifically computer-based) technological revolution in business data processing.14
Three particularly important and very different works of historical scholarship have been published to date on early corporate adoption of computers in the United States: Thomas Haigh’s article on the first five years of the “electronic revolution” in business data processing, JoAnne Yates’s book on punch card tabulation and computer data processing in the life insurance industry, and James W. Cortada’s trilogy surveying computing in more than forty different industries and economic sectors. Haigh highlights the revolutionary rhetoric and provides a meaningful look into the occupations and power struggles in corporations.15 Yates conveys the importance of continuity and gradual migration to new information technology in the life insurance industry and the influence of this trade in helping to shape punch card tabulation and computing.16 Cortada’s three-book survey offers substantial breadth in documenting major currents in the usage of computing over decades in many industries, and government and educational sectors. The focus of all three of these scholars’ works lies outside of the role consulting firms played in the organizational adoption of computers.17
Haigh quotes several 1950s consultants espousing revolutionary rhetoric on computing.18 This accurately conveys a general pro-computer perspective among some consultants in the early computer industry. There were a number of self-identified individual consultants by the mid to late 1950s—some with minimal knowledge but anxious to contribute to and capitalize from a created revolutionary moment in electronics and computing. Other computer consultants, however, began the process of developing true expertise, engaged in exhaustive research and analysis, and became consultancies that client organizations returned to time and again. The remainder of this chapter explores three of those businesses: Arthur Andersen and Company (whose accounting division no longer exists, but whose consulting division is now Accenture, the third-largest IT services firm in the world), Diebold and Associates (which grew throughout the 37-year tenure of its visionary leader John Diebold), and Canning, Sisson, and Associates (which served clients for nearly ten years before Richard Canning left to “scale up” the expertise he sold by entering the technical/managerial advisory publishing business). These three businesses, which were among the first multi-person computer consulting enterprises in the United States, represent the proper beginning of the computer services industry. In all three cases, the themes of constructing expertise (both in terms of building real expertise and in terms of projecting expertise as they learned in a new and rapidly evolving field) come to the fore. The three short case studies presented below also exemplify the computer consulting industry’s role in circulating knowledge and best practices in computing and programming—a defining theme to this day, and one that is evident throughout the computer services industry. The case of Arthur Andersen and Company provides a first glimpse into the somewhat porous and increasingly diminishing boundaries between accounting consulting and managerial consulting, on the one hand, and IT consulting, on the other.
In 1913, Arthur Andersen and Clarence DeLany left the accounting firm Price Waterhouse to purchase the Audit Company of California to form Chicago-based Andersen, DeLany and Company (renamed Arthur Andersen and Company in 1918).19 Since the late nineteenth century, engineers and cost accountants had worked side by side at accounting firms conducting administrative examinations for their clients. Like its industry peers, Arthur Andersen and Company began to perform such services regularly shortly after the launch of the new firm—services that increasingly took the form of modern management consulting.
As historian Christopher McKenna convincingly argues in his book The World’s Newest Profession, cost accounting and the nature of administrative examinations by accounting firms transformed with the rise of “associationalism” in the 1920s as future president Herbert Hoover took control of the Commerce Department. With his engineering mindset, Hoover sought to solve the problem of excessive competition without resorting to heavy government intervention. During Hoover’s tenure as Secretary of Commerce (1921–1928) and as the 31st president of the United States (1929–1933), administrative examinations by cost accountants and engineers provided a conduit for industry-wide sharing of cost information and for conducting managerial audits. Such organizational and strategic audits became the core of modern management consulting. As McKenna relates, in 1933 (early in President Franklin D. Roosevelt’s administration) the Glass-Steagall Act forced accounting firms (and banks) out of this activity and spawned a thriving independent management consulting industry.20
McKenna argues that the federal government took (managerial) consulting away from accounting firms with the Glass-Steagall Act, but enabled a specialized form of it (a gift of sorts), computer consulting, by preventing IBM from entering the field with a 1956 consent decree. While McKenna’s book is a much needed high-quality scholarly history of managerial consulting (largely focused on strategy consulting), his quite brief discussion of computer consulting is distorting. McKenna draws from historian Robert Sobel in relating that IBM sales and service staff would aid with computer installations to make sure everything went right, but fails to recognize the importance of these bundled services IBM provided. He also exaggerates or overemphasizes a counterfactual (if the IBM consent decree with the Justice Department did not exist, no significant independent computer consulting business would have emerged to compete with IBM), and he is off in his chronology. Andersen and Company bundled data processing advisory services before it became a focused Administrative Services Division, and IBM bundled software and services for decades before it became a major profit center for the firm. An important element of the consulting business was impartial outside advice. Many firms would not have been comfortable securing all their advice and guidance from one source—IBM, the leading hardware firm (or from any of IBM’s mainframe-producing competitors). The notion that IBM could have controlled all or even most computer services (and especially consulting advisory services) seems dubious. IBM is the largest computer services company in the world today and is fully committed to this area as its primary business, yet its share of the global market is less than 6 percent.21
In the early postwar years, a decade before the IBM consent decree, Arthur Andersen and Company expanded its Administrative Accounting Division, which specialized in various administrative services, including punch card tabulation. In the 1930s and the early 1940s many of its professionals had learned about business data processing technology from working with punch card tabulation systems at client firms. More significantly, one technology-focused manager and two future Andersen employees who had gained deep expertise in state-of-the-art punch card systems in the military during the war led to Andersen and Company’s revitalizing administrative advisory work for clients on punch card tabulation systems and, later, digital computers.
Leonard Spacek, who upon Arthur Andersen’s death in 1947 succeeded him as managing partner, began his tenure with Andersen and Company in 1928.22 During his 45 years at the firm, Spacek served countless clients, but estimated he did fewer than ten accounting audits.23 His specialty was administrative services, a division of Andersen he expanded after gaining leadership of the company. Shortly after World War II, Spacek hired John Higgins and Joseph Glickauf. Higgins had worked at Andersen and Company previously before joining IBM and later serving in the Navy during the war (specializing in management of mechanization projects). Glickauf joined the Navy in 1942 and worked on developing more efficient payroll and inventory systems at the Naval Bureau of Supplies and Accounts in Cincinnati. Spacek hired these two naval punch card tabulation system specialists not for accounting (Glickauf, in particular, had little knowledge of accounting), but to advise clients on the area of expertise they had gained during the war: data processing systems.24
In 1950 Spacek met with Willis Gale, the chief executive of Andersen’s largest client, Commonwealth Edison. Commonwealth Edison was not merely an accounting client; it also retained Andersen and Company’s advisory services, the Administrative Accounting Division, to improve management systems to track its extensive property records. At the meeting, Gale, who had a deep technical interest and curiosity, asked Spacek about possible digital computer applications for data processing. Spacek assured Gale he would look into that possibility right away and assigned Joe Glickauf to the task. Glickauf met with J. Presper Eckert and John Mauchly and received a demonstration of the UNIVAC later that year. In 1951 Glickauf and other Andersen employees built a small digital computer, dubbed the Arthur Andersen Demonstration Computer (AADEC), to advance their digital computing knowledge and skills. Spacek, Glickauf, and Higgins selected five of the most technically astute Andersen employees to study digital computing—one of whom, Joe Carrico, spent six months with Eckert and Mauchly (both of whom had become friends with Spacek). By 1952 the Administrative Accounting Division was renamed Administrative Services Division to more accurately represent its activities (primarily data processing system advisory services).25
On March 27, 1953, General Electric executives, having heard about the commitment Arthur Andersen and Company had made to researching digital computer applications, hired the company for a $64,000 feasibility study for a UNIVAC I installation to run payroll processing and other systems for GE’s new Appliance Park facility in Louisville. Later that year, in anticipation of the UNIVAC that would arrive at Appliance Park in a matter of months, Andersen and Company’s Administrative Services Division conducted its first “Programming and Installation School” (held at the downtown Chicago campus of Northwestern University) for Carrico and other pioneers to teach 26 new hires or assignees to the Administrative Services Division.26
John Higgins and Joseph Glickauf, the leaders of Andersen and Company’s Administration Services Division, published an article titled “Electronics Down to Earth” in the March-April 1954 issue of the Harvard Business Review.27 It highlighted the General Electric Appliance Division’s planned UNIVAC installation, but more broadly it raised questions for businesses contemplating acquiring computers for business data processing. As the title suggests, Higgins and Glickauf tried to convey balance in the face of the grandiose, revolutionary rhetoric that promised the computer as the cure for all business ills. In the article they mentioned a second client for which Andersen and Company had conducted a computer feasibility study—a firm that was significantly smaller than GE. That unidentified firm had 5,000 employees and a complex payroll because of its excessively high number of job and benefit classifications. Andersen and Company’s suggestion to the management of the firm was to reduce unneeded complexity by having fewer classifications so as to be able to run payroll economically using their existing pre-computer punch card tabulation equipment. While Higgins and Glickauf’s article is optimistic on the future of digital computers for business data processing, it focuses on the importance of deep analysis in finding appropriate solutions for individual organizations and their circumstances. Higgins and Glickauf also argued that the growing preoccupation with high-speed printing was unwarranted, as most organizations could not benefit significantly enough to offset the higher costs of the equipment. Virtually all computer services consulting companies sought to follow the lead of Higgins and Glickauf and to publish business advisory articles and books early and often. Though some resorted to far less realistic assessments by blindly promoting the coming revolution for business (probably reasoning that it would result in more consulting revenue), others, Higgins and Glickauf among them, provided thorough, deeply reasoned, and customized research-based studies.
The manager of the Business Procedures Section of General Electric’s Louisville plant, Roddy Osborn, was in charge of the company’s UNIVAC installation project. Not missing an opportunity to promote GE, he published a company-aggrandizing article titled “GE and UNIVAC: Harnessing the High-Speed Computer” in the Harvard Business Review. “[W]hile scientists and engineers have been wide-awake in making progress with these remarkable tools [digital computers],” Osborn stated early in the article, “business, like Rip Van Winkle, has been asleep. GE’s installation of a UNIVAC may be business’s first blink.”28
As became commonplace in the years and decades that followed, the outsourced advisory, training, and system-building services were downplayed by the client, saving brief mention of Andersen and Company for the next-to-last page of the nine-page article.29 Though what Osborn did write went further than many future clients in publicly recognizing computer services contractors, he commented that “because of the pioneering aspect of the program, we found it wise to employ an independent management consultant firm … Arthur Andersen … experienced in computer logic and developments.”30
More than many other early computer consulting firms that focused on advisory services, Andersen and Company quickly contracted to provide computer programming services as well. It contributed critical programming work for the GE payroll application in 1954, and it probably was the first firm in the United States to engage in contract work for computer programming services. Running into problems, the Andersen and Company team had to engage in reprogramming the entire payroll system, which Glickauf promised that Andersen and Company would do without additional charges. Roddy Osborn and other executives at Appliance Park understood the challenges of pioneering in the new field of applications programming, wanted to be fair to Andersen and Company, and agreed to pay for the reprogramming on a cost-plus basis.31
The high-profile GE project helped establish Andersen and Company’s computer consulting expertise and led to a rapid succession of additional clients in the mid 1950s, including computer installations for the International Shoe Company in St. Louis and for Andersen’s longtime client Commonwealth Edison. The latter contract demonstrated how Andersen and Company successfully practiced account control to transition clients from pre-computing administrative services to computer services. Andersen’s Administrative Services advised Commonwealth Edison to use IBM’s first entry into the electronic digital computing market, the IBM 701, and completed this installation at the start of 1956. Advising and installing successive IBM models became commonplace for Andersen in the years that followed—with Commonwealth Edison, Field Enterprises, and others.32 Beginning with the installation at GE’s Appliance Park, John Higgins established a standard practice of recording, indexing, and sharing computer consulting experiences throughout Andersen’s Administrative Services Division. This facilitated and advanced a culture of circulating knowledge, which greatly aided many clients and proved important to the division’s success.
By 1960, Andersen and Company’s Administrative Services Division had grown to 150 employees. Around that time, the division also began providing large-scale systems integration services to commercial clients, and by the mid 1960s it had extended such services to the federal government. Throughout the 1960s, the 1970s, and the 1980s the enterprise grew rapidly. Andersen Consulting Services (as the computer services division was called by the late 1980s) came to exceed the company’s accounting business in profitability. In 2000 it was split from the company, and it soon adopted the name Accenture. By 2001, Accenture, with 65,000 employees, was the largest exclusively consulting firm in the world.33
Arthur Andersen and Company was unusual in the early computer consulting services industry. Far more than its peers in accounting services and management consulting, it devoted substantial resources to quickly building and expanding other computer services offerings (particularly programming services). Most pioneering computer consulting operations began as startup firms with one or several associates launching and slowly growing enterprises that focused on advisory services. (These businesses either did not venture into programming services, or did so far more gradually than Andersen and Company.) One such firm—(John) Diebold and Associates, later Diebold Group, Inc.—stood out for helping to launch the consulting services segment of the industry, and for its charismatic and visionary leader.
Figure 1.2 John Diebold, circa late 1960s. Courtesy of Charles Babbage Institute, University of Minnesota.
John Diebold was born in 1926 in Weehawken, New Jersey. After graduating from Weehawken High School, the 18-year-old Diebold entered Swarthmore College, but soon took a hiatus to enter the US Merchant Marine Academy, where he developed a fascination with machines and particularly with radar-controlled automatic tracking and firing systems. In 1946 he returned to Swarthmore, where he got a BS degree in engineering. Upon graduating, he entered the Harvard Business School’s MBA program. At HBS he worked closely with George Doriot, a pioneering venture capitalist who had co-founded American Research and Development Corporation (ARDC) in 1946.34
Both Diebold’s experience at the Merchant Marine Academy and his studies at the Harvard Business School contributed to his growing fascination with automatic machinery. A close faculty colleague of Doriot’s at HBS, Curtis Tarr, proved influential. In 1950, Tarr served as faculty advisor to Diebold’s project to research and write a report titled “Making the Automatic Factory a Reality.”35 Diebold’s fascination with digital computers grew rapidly, and he studied the work and writings of John von Neumann and Norbert Weiner, both of whom he contacted and met with in the early 1950s.36 Diebold would go on to substantially revise and expand his HBS report into his soon-to-be-famous 1952 book Automation: The Advent of the Automatic Factory.37
In 1951, Diebold completed his MBA program. He then sought employment as a management consultant and was hired by a small Chicago consulting firm led by E. O. Griffenhagen. Diebold incorporated his factory automation studies into certain consulting assignments and was successful working for Griffenhagen. Nevertheless, within several years Diebold wanted the freedom of his own consulting operation. In 1954 he returned to Weehawken to found a consulting enterprise.38
Highlighting the tremendous importance of publishing to establish name recognition and conveying expertise in consulting, Diebold published frequently throughout in his career. Diebold did not invent the term “automation” with his 1952 book, but that book and his subsequent publications and work did much to advance both the use of the new term and computer-based automating efforts of various kinds.39 The subtitle of Diebold’s book suggests that it is focused solely on automating production machinery and processes within factories, and indeed the first two thirds of the book concentrates on this topic. The final third, however, considers the future of automation of information in offices and other settings and how it might impact organizations and society. Though Diebold clearly advocated the use of computer applications for the betterment of business, government, and society, he was not blindly optimistic; he recognized the importance of deep organization-level and industry-specific analysis. Intensive research was the guiding force of John Diebold and Associates, the computer consulting business he founded in Weehawken. (Later it relocated to New York City.)
Most of the small, privately owned computer consulting firms that went out of existence twenty years ago left little or no trace for researchers interested in their past activities. In contrast, John Diebold left behind a rich archive (housed at the Charles Babbage Institute) detailing his firm’s work—a complete set of more than 500 consulting reports the company prepared for corporate and organizational clients throughout the firm’s 37-year existence.40
Diebold and Associates contracted with their first clients in 1956. That attracting the first customers proved difficult is understandable insofar as the very first movers hired the well-established Arthur Andersen and Company in 1954 and 1955. Thereafter, the growing attention to Diebold’s book helped with name recognition and attracting clients. Table 1.1 lists the firms that Diebold and Associates served from the mid 1950s to 1960. From the early 1960s on, the company expanded steadily and produced far more reports each year.41
Table 1.1 Companies and organizational clients of Diebold and Associates, 1956–1960.
| Alwac Corporation |
| Alweg Monorail |
| American Hospital Association |
| AMF Pinspotters, Inc. |
| Arco Auto Carriers |
| Bear, Stearns & Company |
| Boeing Airplane Company |
| Bulova Watch Company |
| Carl M. Rhodes & Company |
| Coopers Inc. |
| Department of Highways, Ontario |
| Federal Electric Corporation |
| General Electric Company |
| Gladding McBean & Company |
| Illinois Agricultural Association |
| Kellogg Switchboard & Supply Company |
| Kleber-Colombes |
| Mead Johnson & Co. |
| Owens-Illinois Glass Company |
| Pepperell Manufacturing Company |
| Radio Corporation of America |
| Remington Rand |
| Shuron Optical Company |
| Socony-Mobil Oil Company |
| Stromberg-Carson |
| Swiss Bank Corporation |
| Teleautograph Corporation |
| The Peoples Gas Light and Coke Company |
| The State Insurance Fund |
| Western Virginia Pulp Paper Company |
| Westinghouse Electric Corporation |
Table 1.1 illustrates the diversity of Diebold and Associates’ early clients. Most of the studies that the firm conducted were contract work for corporate or organizational clients. By 1960 the firm had already consulted on potential acquisitions and implementations of computing and software systems at American corporate giants, smaller enterprises in the US and overseas, governments, public utilities, hospitals, and other types of organizations. Applications ranged from business data processing and competitor research to factory automation and developing aerospace control systems. Diebold and Associates provided expert advice not only on technology and systems, but also on a myriad of organizational and managerial issues involved with IT transformations. A typical consulting project lasted from six weeks to six months and yielded a report of 80–250 pages presenting extensive research findings.42 Among Diebold’s clients between 1956 and 1959 were Alwac, Mead Johnson, and the American Hospital Association.
Alwac is not a widely recognized name in computing, despite the involvement of the highly successful Swedish entrepreneur Axel Lennart Wenner-Gren.43 Wenner-Gren had made his early fortune in the vacuum cleaner and refrigerator industries. In 1952 he became involved in computing, partnering with an ex-Northrop Aircraft engineer to form Alwac in Redondo Beach, California. Like several of other mainframe digital computer companies that formed in the Los Angeles metropolitan area (a hotbed of aerospace and defense contracting), the firm struggled as a business. In contrast to several computer industry leaders—that transitioned from pre-computer office machine specialists to become dominant in the computer industry, or startups fortunate enough to be acquired by such firms—most new ventures in digital computing of the early 1950s had periods of growth, but their existence was always precarious.
As Alwac struggled to contain costs in the mid and late 1950s, it brought in Diebold and Associates to do a thorough analysis of the firm and its mechanisms, competitive environment, and product and market opportunities. Diebold and Associates found a firm with few controls in place that was not basing its strategy on sound evaluations of the market and of its competitors. Diebold consultants put together a series of recommendations for controlling costs and improving the organization of the firm. Alwac had two programming groups that did not coordinate or communicate effectively. The programmers had little understanding of the internal circuitry of the ALWAC I computer, the company’s initial product. Diebold consultants developed a plan to put the programmers in a single large programming room and to educate them on the requisite computer circuitry. That a computer startup in the 1950s would have difficulty organizing programming efforts is understandable—programming would prove a major challenge for even the largest computer firms in the 1950s and the 1960s. For instance, IBM experienced delays and challenges with FORTRAN, OS/360, SABRE, and other major internal and external programming projects. Diebold and Associates brought the combined wisdom of studying and learning from a number of programming efforts by the time they consulted for Alwac in the late 1950s.44
Despite sound advice from Diebold and Associates, Alwac continued to suffer severe cash-flow problems and ultimately could not compete effectively with the office machine and electronics leaders (IBM, NCR, Remington/Sperry Rand, GE, RCA, and Burroughs) that, along with two well-funded and well-organized startups from 1957, Control Data Corporation and Digital Equipment Corporation, would dominate the computer industry in the 1960s. Nonetheless, it is telling that Alwac contracted with Diebold and Associates to work continuously for two years to examine and advise on every aspect of the firm: engineering, design, product development, programming, marketing, and accounting. Much of the expertise that enabled Diebold and Associates to aid Alwac, as well as more successful computer industry participants (Burroughs, IBM, RCA, GE), came out of its experience at teaching and learning from office and factory automation users and emerging users, who made up the bulk of its clientele.45
In 1895, Edward Mead Johnson (one of the three brothers who had founded Johnson & Johnson in 1886) launched a side business, American Ferment Company, to produce and sell digestive aids. Two years later, he left Johnson & Johnson to focus on American Ferment, which was renamed Mead Johnson & Company in 1905. The company became a significant producer of infant formula, diet aids, and other products. In 1957, it contracted with Diebold and Associates to do a preliminary study of possibilities for administrative data processing.46
Although Diebold and Associates reported preliminary assessments of how an electronic digital computer system could help schedule production and minimize total costs (material, purchasing, manufacturing, and distribution), achieve faster and more accurate customer invoicing, provide current sales analysis reports, and assist the Research Division with the development of new products, the true focus of the six-week study was on the existing Tabulating Department and its operations. That department’s technology consisted of pre-computer IBM punch card tabulation equipment. Diebold consultants determined that the firm had major organizational and managerial problems. Although Mead Johnson & Company had been hiring more and more employees and had been increasing its output for decades, the Diebold consultants argued its Tabulating Department was failing to meet the company’s needs. They found that the department’s manager lacked the ability that was needed to exercise control over personnel and equipment. They noted that he failed to educate users of his department’s services, and also that he promised to complete reports on impossibly short time schedules and then failed to deliver them. At the same time, this manager’s supervisors missed opportunities for the department to serve the corporation on account of a short-sighted personnel policy, and they failed to require coordination between the Tabulating Department and other departments. The Tabulating Department’s manager was not being invited to appropriate meetings in other areas of the firm and was being kept in the dark about initiatives that would require his department to provide new services in the future.47
Computer technology came at a high cost, and organizations had to be creative, cooperative, and proactive in seeking to fully take advantage of new systems. The skill sets of managers and operators in tabulating departments often were insufficient for a larger-scale computerized operation. Diebold consultants recognized and recorded the inadequate technical understanding of Mead Johnson’s Tabulating Department manager, but put far greater emphasis on the organizational and managerial skill set a new manager would need to fully take advantage of computing within the company, and the way different sectors and divisions of the corporation would need to cooperate in new ways. Ultimately, Diebold consultants provided not only technical evaluations, but also thoughtful recommendations for recruiting, training, promotion, and knowledge sharing to facilitate a more competent Tabulating Department regardless of whether Mead Johnson invested in computer technology.48 More generally, Diebold and Associates’ evaluations assessed existing and potential new technology within Mead Johnson’s managerial, organizational, and personnel contexts. Mead Johnson & Company’s management found the study so helpful that the firm contracted with Diebold for three additional studies over the succeeding decade.49
The American Hospital Association (AHA) was founded in 1898 to promote good practices related to hospital health care and health administration. In 1958 it hired Diebold and Associates to analyze and make recommendations for Baylor University Hospital (in Dallas)—as a test bed—in order to establish standards and best practices in health administration and data processing that could be utilized at many other (AHA-member) hospitals. Diebold consultants analyzed current practices and evaluated options and opportunities for computing applications to the following functions: accounts receivable, accounts payable, general accounting, payroll, inventory control, purchasing, and medical records and statistics. While the research focused on Baylor University Hospital, Diebold and Associates gathered extensive data on several other hospitals for comparative purposes, including Memorial Hospital for Cancer in New York and Barnes Hospital in St. Louis.50
Diebold consultants provided a complete plan for how a data processing infrastructure based on digital computers could be implemented at Baylor University Hospital. The consultants performed an extensive cost analysis of the existing record-keeping and retrieval mechanism, and projected costs for a computerized record system. They concluded that a general-purpose computer with large random-access memory and relatively low arithmetic and logic speeds would suffice for processing a hospital’s records. They stipulated that such a system wasn’t likely to bring savings if installed right away (early 1959), but that cost trajectories on systems and memory were such that computers could probably bring savings to most large hospitals, including Baylor University Hospital, within three to five years. (Computing costs would continue to decline relative to processing power and memory, and existing clerical costs would continue to rise.) Diebold and Associates provided a highly detailed plan for utilizing punched card tabulation and digital computer systems to organize and use data.51
In the late 1950s, the field of medical computing was gaining attention (focused on patient records) because of the work of Robert Ledley and Lee Lusted, radiologists who published extensively on how computing could benefit diagnosis. Their ideas would not be realized until ten years later, when the cardiologist Homer Warner instituted some of them at Latter Day Saints (LDS) Hospital in Salt Lake City. LDS Hospital was acquired by Intermountain, which remains at the forefront of the still somewhat underdeveloped fields of computerized diagnostic tools.52 On the other hand, computer applications for hospital administration grew significantly in the 1960s, in the 1970s, and afterward. Diebold and Associates’ recommendations, and the associated test bed system at Baylor University Hospital, contributed to this phenomenon.
Over time, Diebold’s firm supplemented client-based work with producing and selling industry-specific analysis reports, or position papers on IT trends. This was secondary, as client contracts were always the focus of the firm. Many of Diebold’s clients became repeat customers over years and decades, hiring Diebold and Associates to evaluate and assess new IT systems, new applications, and organizational issues associated with information technological change. Some industries that appear underrepresented on the list of the first thirty clients but which we know were early adopters of computers, such as insurance and banking, had long-established relationships with IBM, Burroughs, and other computer manufacturers, and greater internal expertise with computing at an earlier date. Banking and insurance firms, however, became frequent Diebold clients in the 1960s, 1970s, and 1980s—including Bankers Trust, Carteret Savings Bank, Dollar Savings Bank, Bear Stearns, Blue Cross and Blue Shield, American Insurance Services Group, American International Group, Central National Insurance, and Travelers. From natural gas, oil, aerospace, investment, agribusiness, and pharmaceuticals to paper, paint, electronics, airline, earth-moving machines, and tobacco companies, Diebold’s industrial client base was extremely broad. It even served the trade union Communication Workers of America to help this organization understand the impact of automation on labor issues in their industry. In fact, this organization commissioned nine studies from Diebold between 1964 and 1983 making it one of the firm’s longest-term clients.53 Beginning in the late 1950s, it also served client companies and organizations around the world, with a particularly rich base of Japanese firms and Japanese government organizations—including Mitsubishi, Japan Information Processing Development Center, Japan Trade Center, and the Japanese Ministry of Industry and Trade. Likewise, the firm consulted for many different local governments, and federal departments and agencies. The mix of industries and government, and geographic diversity of clients, helped to insulate Diebold’s firm from economic downturns in particular regions and economic sectors.54
To judge from the number of client contracts, completed reports, the length and depth of studies, and the fact that most projects involved sending a small team, Diebold and Associates must have had at least a dozen consultants on staff by 1960, and many dozens by the late 1960s (by that time under the new name, Diebold Group, to reflect its larger size). The number grew steadily in succeeding decades. At its peak, in the 1980s, the firm had more than 300 employees.55 Soon after launching the enterprise, John Diebold became more of a manager of consultants and programming specialists than a consultant himself.
John Diebold, as well as many of his consultants on staff, published frequently. After the publication of Diebold’s highly popular book Automation, publications had a major effect on the growth of his business. Not only was publishing a free or inexpensive marketing tool, it was a particularly effective one because articles and books in respected journals and published by distinguished presses conveyed expertise—exactly what Diebold’s firm sold. Diebold and Associates/Diebold Group conducted dozens of consulting projects for clients in the second half of the 1950s, and more than 100 in each of the three succeeding decades. In all it served more than 300 unique client organizations, introducing many to digital computing, new applications, better organizational processes and structures, and more effective and efficient use of information technology. In 1991 John Diebold sold Diebold Group, Inc. to Daimler-Benz.56
Though small compared to Andersen and Company’s consulting operation, Diebold and Associates was a substantially larger enterprise than many emerging computer consultant firms in the early industry—some of which were sole proprietors or partnerships. Perhaps the earliest partnership in computer consulting services was Canning, Sisson, and Associates, which began in 1954, the same year as Diebold and Associates.57 Canning, Sisson, and Associates—founded by Richard Canning and Roger Sisson—had a somewhat different business model than either Andersen and Company’s Administrative Services Division or Diebold’s firm. A core activity of Canning and Sisson’s company was spreading expertise in computer systems and applications by holding fee-based educational seminars.
Richard Canning, who had worked on a new radar bombing set during World War II while serving overseas, joined IBM immediately after the war and was assigned to production engineering. Less interested in the production of IBM machines than in their industrial application, he soon took an IBM course on manufacturing control at the firm’s facility in Endicott, New York. The rigidity of IBM and the experience of raising a family in upstate New York, where winters were harsh, led Canning to seek greener pastures with a smaller and more flexible firm in Southern California. Canning would be the first of many who gained experience at IBM, later to leave for computer consulting services or other types of careers in computer services.58
After writing some letters to Southern California firms, Canning was offered and accepted a position with Electronic Engineering Company in February 1950 (located in Santa Ana). About two years later, to augment his income and pursue an interest in computer applications to manufacturing that he had been unable to at IBM, Canning joined a research group as part of the Management Sciences Research Project (funded by the Office of Naval Research) at the University of California at Los Angeles, where he had the title Associate Research Engineer. The group was focused on applying computers and operations research to aircraft manufacturing, and on advising the ONR’s Logistics Branch. Roger Sisson, who had a background in electrical engineering and had worked at National Cash Register previously, joined the UCLA project on the same basis. Soon thereafter, in fall 1954, Canning and Sisson decided to form a computer consulting firm together. Its name was Canning, Sisson, and Associates.59
Back in 1953, before the installation of the very first data processing computer to be used in business in the United States (a UNIVAC at GE’s Appliance Park), Canning began teaching a night school course on computer applications for business in Los Angeles. He also taught a one-week condensed version of the course at the University of Chicago. There was no textbook on the subject, so Canning began to write one. Ultimately it became Electronic Data Processing for Business and Industry, a seminal work in the early digital computer business applications field, published in 1956.60 The latter portion of Diebold’s 1952 book Automation had suggested the possibility of widespread computer applications in business, but Canning produced the first high-quality textbook detailing state-of-the-art knowledge about applying computers to a range of such applications—from department store inventory control and airline reservations to managerial decision making. Context and understanding provided by Canning’s early participation in the pioneering ONR study at UCLA enabled him to write this deeply researched text.61 What stands out about the volume is that it does not present a wildly optimistic utopian vision; rather, it presents a sober in-depth evaluation of business processes, operations research, and existing equipment and methods of applications. Canning notes that the shift to electronics and digital computing did not have to be wholesale, and that in most instances a wholesale shift was unwise. He emphasizes that firms and organizations can gradually incorporate electronics equipment to complement and improve functions and run hybrid electromechanical and electronics data processing systems. He discusses ways to make optimal use of digital computers, but argues that firms and organizations could (and often should) make gradual transitions that advance operations while containing costs.62 This is also true of Canning’s second book, Installing Electronic Data Processing Systems (1957), in which he offers advice on opportunities and challenges by drawing on specific cases of installations, hiring and organizing of data processing personnel, general programming, programming computer runs, costs of operations, and other topics.63
From the start, Canning and Sisson decided to avoid any real or perceived conflict of interest by refusing to simultaneously consult for multiple companies in the same industry. Both men had started out in operations research and computer applications to aircraft production control and were optimally situated (in terms of knowledge and geography) to serve aerospace firms. Limiting themselves to one aerospace firm at a time, they quickly conducted research and developed broader expertise to provide quality advisory services on electronics and computer data processing applications for other industries.
Knowing the substantial tuition that University of Chicago charged for his popular week-long course, Canning, the more talented educator of the two, began teaching one-week fee-based seminars on computer applications to business data processing and other industrial applications in Los Angeles, in New York, and in Chicago. Sisson would sometimes co-teach the Los Angeles seminars with Canning, but in order to keep travel costs down he seldom traveled with Canning for the New York and Chicago seminars. The five-day events carried a $200 fee, which included lunches, and were held in luxury hotels. Canning and Sisson downplayed their roles as consultants at these events, so attendees did not feel they were paying a substantial fee just to get a sales pitch—and when asked about consulting, they often recommended major accounting and management consulting firms such as Andersen and Company, Coopers and Lybrand, and McKinsey and Company. Nonetheless, attendees often suggested getting together for dinner or drinks, which led to site visits to companies, and ultimately significant consulting contracts for Canning and Sisson, particularly on the East Coast. One such client was the Associated Merchandizing Corporation, a purchasing association for member US department stores. Recording, processing, and analyzing information quickly was (and is) critical to department stores’ purchasing and cost containment. In the late 1950s, AMC decided to move forward with a project in which AMC members would support a single computer project to analyze information and better understand how computers could be applied, using one store (a Higbee’s department store in Cleveland) as a test site. Canning and Sisson advised on hardware, on accounts payable, on payroll, and on other software applications for the test site.64
Over time, publications—Canning’s first book, and both Canning’s and Sisson’s ongoing publication of articles and books—were important to the continuing success of the business. As with all consulting firms, getting off the ground with initial clients was the most difficult part. Canning’s and Sisson’s work on the ONR project, their publications, and holding high-level seminars all brought name recognition, leading to early clients—and successful projects produced invaluable word-of-mouth marketing when data processing managers came together at professional meetings such as the National Machine Accountants Association/Data Processing Management Association (NMAA/DPMA), or other events and venues.65
For Canning, Sisson, and Associates, in contrast with most computer consultants, publishing was not merely a critical tool to display expertise and establish a brand; it was also a business. To be successful, consultants had to keep abreast of current literature on computer applications in industry. This included the occasional article on the topic in well-known business publications such as the Harvard Business Review, but also many trade journals for individual industries as well as technical publications. Roger Sisson reasoned that data processing managers (potential clients) had to do the same. With this in mind, Sisson and Canning decided to launch Data Processing Digest as a monthly subscription-based publication to synthesize useful information for busy managers. Sisson had supervised Margaret Milligan, a technical writer at National Cash Register, who was hired away from NCR by Canning and Sisson to write the content for and oversee Data Processing Digest. She requested and obtained blanket permission to synthesize or summarize articles from most of the relevant publishers, and would spend endless hours in the UCLA library looking for appropriate articles to summarize. The first issue (April 1955) consisted of ten pages. While Canning and Sisson would provide some advice and occasionally write, Milligan did most of the writing, editing, formatting, and managing of the publication. Sending out subscription forms to 3,000 contact addresses Canning and Sisson had assembled, they quickly secured 300 annual subscribers—a base that would grow steadily. Milligan networked within the emerging computer community and was often the only woman at meetings of the Los Angeles chapter of the Association for Computing Machinery.66 As with data processing managers, emerging computer consultants, and others, Canning and Sisson benefited from the publication as a time-efficient tool to stay current on the literature and to make wise decisions on what articles to read in full.
Canning and Sisson were well aware of Diebold and Associates in the mid 1950s, and Sisson visited Diebold once in his New York office, but the two computer consultancies generally did not compete for contracts. Canning and Sisson usually sought shorter-term contracting assignments and lacked the geographical scope of Diebold’s firm. From the beginning, Diebold aimed to build a steadily growing consulting corporation, while Canning and Sisson were focused on running a partnership where they were the primary consultants. Over time, however, Canning and Sisson differed on their vision for the future of the company. Roger Sisson wanted to expand into programming services and become a manager of a broader-based computer consulting company, along the lines of Diebold and Associates or Arthur Andersen and Company’s Administrative Services Division. Richard Canning, in contrast, was more content with advisory consulting, teaching seminars, and operating a very small firm. With these differences, the business partners decided to part ways in 1958. Canning continued to consult and hold seminars, and in time, sought to leverage his expertise by starting a new publication, while Sisson abandoned consulting and joined the Ford Motor Company as an aeronautical engineer.67 In 1961, Canning and Sisson sold Data Processing Digest to Milligan, who expanded it to include commissioned book reviews and articles and published it for another 25 years.68
Data Processing Digest alerted Canning to the potential business of publishing on computer applications, and in September 1962 he had a career-changing idea. Whereas Data Processing Digest gave a quick synopsis of many of the relevant articles on computer-based business data processing, Canning envisioned a new publication to focus on original content and to offer greater depth. Most articles on business data processing were quite short, usually no longer than a few pages. Canning’s plan was to launch a publication, to be named EDP Analyzer, in which he would write a single in-depth article of 10–16 pages concentrating on a significant theme in the field for each monthly issue. The subscriber base was primarily companies, not individuals, and EDP Analyzer was priced as an expert journal for companies/institutions at $36 (approximately $280 in today’s dollars) a year. With the boldness of a true entrepreneur, Canning invested all the money he had—$4,000—in direct mail marketing for the first issue (and offered a money-back guarantee to initial subscribers). He wrote all the issues from 1963 through 1977, sharing duties with his daughter until 1987, when publication ceased.
Throughout 1963, Canning continued consulting, but subsequently he devoted full time to the journal. In his years publishing EDP Analyzer, Canning traveled extensively to spark ideas for future issues’ themes. His consulting days had provided him with a rich network of friends and contacts. In that sense, his post-consulting publishing career mirrored many aspects of consulting—traveling extensively, meeting many data processing and computing professionals, engaging in research, and circulating knowledge. EDP Analyzer leveraged this dissemination of expertise, as the subscriber base grew substantially and eventually reached a peak of 10,000—a remarkable feat insofar as many organizations had only one subscription and shared or illegally photocopied the issues for all members of the data processing staff.69 The same down-to-earth advice Canning gave as a consultant—emphasizing deep analysis and cost-effective evolutionary rather than revolutionary change—radiated throughout the pages of EDP Analyzer.
The computer consulting sector was the origin of the computer services industry, and Arthur Andersen and Company’s Administrative Services Division was probably the first computer consulting enterprise. Most computer consulting companies were small—indeed, many were one- or two-person operations, like Canning, Sisson, and Associates. These sole proprietorships, partnerships, or very small corporations tended to leave little or no documentary record of their existence behind, despite being enormously influential in aggregate in providing expertise in the emerging field of computer-based business data processing.
John Diebold almost singlehandedly popularized the term “automation” for manufacturing and business data processing, and in 1954 he founded the most significant startup computer consulting corporation. Ultimately computer consultants sold their expertise and sold their time. From the standpoint of individual consultants, the time limitation (of perhaps 2,000 or 2,500 billable hours a year) is what led Andersen and Company to expand into programming services and become a large enterprise, led John Diebold to add consultants and programming professionals and become a manager of a midsized firm, and led Richard Canning to leverage his knowledge and expertise by entering the publishing business.