What Is a Business Information Ecosystem?

There is no hard consensus among business historians concerning the answer to this question, so far. Those who have pondered it have relied largely on three inspirations: the biological analogy of how animals, plants, earth, and climate interact with one another and thus are codependent for their success;Footnote ⁶ sociologists exploring how societies function, describing interactions among their human constituents, also codependent for the overall survival of its members;Footnote ⁷ and increasingly the role of corporate cultures serving as ecosystems.Footnote ⁸ All three approaches share a few common generalized features. Each participant—say an employee or company within an industry or market—is affected by the activities of the others, such as what competitors do in their markets. Each has a working knowledge of how other participants behave, generally, and develops (or has preconceived strategies) for responding to these in ways that optimize its own well-being. In business one speaks of this largely in the language of tactics and strategies. In each ecology—environment—there is broad, if unstated, awareness of important realities, such as the seasons, which can also be characterized as annual events within a firm or industry (e.g., fall planning exercises or major industry conferences); social and religious norms to be accounted for in commercial behavior and in business laws and regulations; and shared beliefs in managerial behavior and the implementation of “best practices,” such as what people are taught in business schools or on the job. All can be applied to understand how a corporation works.

Historians have long recognized, for example, that corporations function within a market, a national economy, and within the realities of laws and regulations. They just never called it an ecosystem, with the result that the more traditional view is often one-sided, whereas an ecological perspective holds out the promise of a more multidimensional view of a corporation’s behavior. Adding to the mix a sociological dimension—what we call corporate culture—enriches our understanding of how an enterprise operates. Historians have increasingly seen the value of considering this facet of corporate life. For decades, IBM employees and historians of the firm have long argued that the company’s culture was the primary influence on its behavior over the past thirteen decades. What they were saluting was yet another facet of a business ecosystem.Footnote ⁹

There is another component, drawn from studies in biology, sociology, information technologies, and communications since the 1930s, which has come to be known as feedback loops, most widely brought to our attention in the 1940s by such researchers as Norbert Wiener, a mathematician at MIT, and Claude E. Shannon, a Bell Labs researcher focusing on the transmission of information.Footnote ¹⁰ What they and others contributed, and that Yates, Beniger, and other historians picked up on, was that the entire edifice of an ecosystem depended on the collection, use, and transmission of information by all the inhabitants of an environment, be it in a market, company, jungle, family, or a society simply trying to maintain family relations through reliable long-distance telephone conversations. It is increasingly becoming evident across multiple academic disciplines that the role of information is the glue that makes possible the other behaviors normally attributed to an ecosystem. Here we give it a label—information ecosystem—to call attention to it, relying on the biological metaphor.

Here, an information ecosystem consists of all the information that participants in a company, market, or society are exposed to, rely upon, and are affected by as they go about their activities. A business also has its own information ecosystem—we normally think of it as part of corporate culture—but it is the factual infrastructure that holds up the entire edifice of a corporation. It can be identified within the confines of an enterprise. First, the ephemera of an information ecosystem includes paper objects, such as annual reports, memorandum, manuals, sales plans, blueprints, personnel manuals, marketing literature, reports, documents articulating personnel and marketing behaviors, catalogs, even posters extolling corporate values, memoirs and histories, slide presentations, and CEO quotes carved or painted on the walls of company buildings. It is a diverse and massive quantity of evidence.

Second, there are verbal components too, such as speeches and conversations delivering consistent messages about values, behavior, plans, and objectives. These transmit information, often consistent with those found in a company’s paper records. IBM’s case is a clear example of the two entwined.

Third, there are the actual specific pieces of information needed to conduct business. In IBM’s case, these include descriptions of specific products (e.g., features and costs); systems for how sales personnel will be paid (e.g., certain dollar bonuses for selling specific goods within a prescribed time); and process documentation that explains how to write a specific piece of software or to assemble a specific product (e.g., tabulator, computer, terminal). Then there is all the information that historians always documented, such as accounting, finance, market share, and competitive responses. There is also the mix of verbal communications entwined with the documentary records that brings to life what is done within a firm that is less formal, perhaps conducted in a bar after work, in a manager’s office when advising an employee, or at home when sharing with a spouse.

When studying an information ecosystem, then, historians need to go beyond the obvious paper ephemera and sociological considerations if they are to broaden their understanding of the extensive role information plays in the life of a corporation, to understand more fully its history. That is before we even observe the most obvious fact that a company such as IBM may essentially be a dominant participant in a market occupied by other information-handling participants that are affected by both their information and IBM’s own. However, the effort is worth it, because historians can arrive at a richer, multidimensional view of a company’s history.Footnote ¹¹

The Argument for Studying IBM as an Information Ecosystem

The specific argument in favor of choosing IBM or other large, long-lived enterprises is straightforward. First, IBM has been studied by business historians, beginning with Robert Sobel in the 1970s and 1980s, next by Stephen Usselman, then by James W. Cortada, Lars Heide, and Richard S. Tedlow, by Chandler in the early 2000s, and most recently by a new generation of scholars exploring IBM’s international role, notably Thomas D. Haigh and Petri Paju, among others, all so convenient for our purpose.Footnote ¹² Second, journalists, economists, and experts on computing technology have also done a great deal of work on IBM.Footnote ¹³ Third, it has long been an important, influential company that, as early as the 1920s, was already emerging as a corporate icon, beginning to be emulated by other enterprises.Footnote ¹⁴ In the second half of the twentieth century, it worked at the center of the world of computing and continued to grow. Today IBM has over 350,000 employees in 175 countries. Over the course of its history, IBM employed nearly one million people (not including business partners) and generated over a trillion dollars in revenues.Footnote ¹⁵

Fourth, its value proposition rested on selling goods and services for the acquisition, use, and management of information by large organizations in the public and private sectors. IBM was in the thick of it all. For decades, the way enterprises thought of how to use information was shaped by IBM’s own worldview. Thus, we can begin to see this company as either at the epicenter of an information-filled world or highly influential within it, at least with respect to how other institutions used tabulators from the 1890s to the end of the 1950s, mainframes from the 1950s to the present, and increasingly today, software security, data analytics, hybrid cloud computing, and artificial intelligence.Footnote ¹⁶

It has a rich corporate archive, which the firm periodically allows historians to access. There are others located at various universities (such as at New York University and the University of Minnesota) with much additional ephemera, and an immense press coverage it purposefully encouraged exists. Dozens of IBM employees have written about the company.Footnote ¹⁷ Another corollary is that issues studied by business historians transcending the experiences of one firm often can include IBM. A good example is the work of Walter A. Friedman, who studied the history of how salesmanship professionalized in the period in which IBM existed, which cites examples (among others) of its activities, augmented by more focused studies of IBM’s sales practices by others.Footnote ¹⁸

So, viewing IBM as an information ecosystem with a complex information infrastructure can begin. We can build on the work of observers who described IBM as collections of technologies and communities identified as users of specific products. These include users of large mainframes in the 1950s–1980s, for example, the role of extensive computing in data-intense industries (e.g., insurance and banking) and in wealthy industrializing economies. Technologists, computer scientists, and employees promoted this worldview. They were able to describe what information they used.Footnote ¹⁹ A second entwined approach views IBM as a social community through the study of its corporate culture, the tact taken almost universally by IBM memoirists. Both approaches contribute to our ability to see IBM as an information ecosystem.

First, IBM as a Social Community

Geert Hofstede, a human resource manager at IBM in the 1960s and 1970s, who after leaving IBM went on to study the role of corporate cultures in multinational corporations, described such cultures as “always a collective phenomenon, because it is at least partially shared with people who live or lived within the same social environment, which is where it was learned.” He added that “culture consists of the unwritten rules of the social game.”Footnote ²⁰ Although Hofstede’s work has been superseded by others, even criticized by students of corporate cultures because of his facile application of cultural stereotyping, his early work remains relevant because it was based entirely on what was happening inside IBM. Views of IBM through the lens of its widely acknowledged strong, indeed pervasive and ubiquitous corporate culture, has long served as a means for understanding IBM. Employees recognized and cultivated a highly articulated perspective since the 1920s. Every memoir written by them paid homage to this culture, as much as they did to describing it. From one generation to another their descriptions were similar regarding its symbols, heroes, rituals, values, and universal homogeneity across nations.Footnote ²¹ That does not mean they universally defended all its features; far from it, as laid off employees testified through their blogs and other comments in the 2000s.Footnote ²² But even they complained about the same issues.

We are, of course, talking about IBM’s fabled, much commented upon, corporate culture. Broad familiarity with it by employees and their customers absolves us from describing it here in detail, but not from explaining why it exists as a way of understanding IBM’s behavior. Historians Philip Scranton and Patrick Fridenson remind us that corporate cultures and how business people work is purposeful, explain why firms are so diverse, and hence why IBM is seen by employees and sociologically oriented historians as unique. In their words, priorities (i.e., culture and work) are chosen, “by fiat, tradition, logic, and emotion; to exclude any of these dimensions is not worthwhile, as a broader scope for thinking and inquiring opens doors we might otherwise miss entirely.”Footnote ²³ IBM employees reiterate their point: why understanding their corporate culture is the primary lens through which to view and understand their company. They, historians, and business management professors privilege this perspective, because the culture aligned with the business strategy of the firm, and it was to the effectiveness of that alignment that observers attributed so much of IBMs success. The value of this perspective has not been lost on those studying other large international corporations.Footnote ²⁴

A company’s culture shapes the behavior and performance of its employees, gives purpose to all, and serves as an extraordinary influence on the destiny of the firm. A successful culture reinforces profitable behavior, which then nurtures a virtuous circle continuing to fortify behavior that leads to more successes in innovation, growth of revenues and profits, and customer satisfaction and their dependence on the firm. That was largely IBM’s experience. The ugly side, also demonstrated by IBM, is that when market circumstances change, such as by the introduction of new technologies and products by competitors, a highly developed culture can prove inflexible, making it difficult for the firm to respond appropriately to existential challenges, as happened in the 1980s.Footnote ²⁵

IBM’s history reinforces several key points about corporate cultures that are useful guides to what to examine in other enterprises. First, there was a corporate culture at IBM that flourished early and that senior and midlevel management nurtured purposefully as an effective way to facilitate its work. That is the main point made by employees. Second, its culture long retained, largely intact for decades, core values known at IBM as the three “Basic Beliefs”: high quality customer service, respect for employees, and excellence in all the company’s endeavors. Third, its culture evolved over nearly one century and not always effectively, as CEO Louis V. Gerstner Jr. noted in his memoirs.Footnote ²⁶ Fourth, employees observed that historians who had not worked in this company never quite “got it right.”Footnote ²⁷ Employees did not either, often describing only facets of it, or describing it in some idealized fashion that experts on corporate culture find superficial or too fractured.Footnote ²⁸

Edgar H. Schein’s model of organizational culture provides a more disciplined approach with which to view this aspect of a company’s way of working. Across five editions of his text, Organizational Culture and Leadership, he defined the components and dimensions of a culture and the role of individuals within it. He argued that corporate cultures have three levels. The first are artifacts of a culture, such as “structures” and “processes,” and observed behavior. The second are the “espoused beliefs and values” such as “ideals, goals, values, aspirations,” in IBM’s case, its Basic Beliefs. Often it was this second level of Schein’s model that received the most extensive attention from employees, historians, and other commentators.Footnote ²⁹ His third, and perhaps most important level, were the basic underlying assumptions about the operation of the firm. He described these as “unconscious, taken-for-granted beliefs and values,” because these determined behavior, perceptions, and thoughts, the way things were really done.Footnote ³⁰ Biographers of key IBM leaders devoted more attention to this feature of IBM’s behavior than other observers.Footnote ³¹ All three levels are perspectives historians are able to address.Footnote ³² They should be direct components of any historical description of a company’s information ecosystem, because these consist of participants, their behaviors, and the reasons for their actions, which offer insights on consequences and the dynamics of life in its ecosystem.

Commentators were, by their attention to IBM, de facto members of its information ecosystem. Their format for describing IBM’s culture proved fairly uniform, too. Everyone embraced the company’s iconic bumper sticker phrase—think—as shorthand for much about the company, and so it has, for them and employees, been a symbolic fixture of the company’s values and behaviors for over a century. It calls to mind two notions—the obvious one that solving a customer’s data processing needs requires cerebral actions, because computers and the work they do are not intuitive and need intellectual energy and tailoring to specific requirements. However, there is a second, less acknowledged, side to think: the act of embracing the wisdom and common sense value of IBM’s way of doing things. This second point is extremely important, which historian Richard S. Tedlow explained: “When Watson said think, he did not mean ‘think for yourself.’ He meant ‘think like me.’” Why? Because, “he positioned himself as the leader with the answers,” and for over forty years successfully transformed a tiny company into a behemoth.Footnote ³³ It proved hard to argue with such success.Footnote ³⁴ It spoke to Schein’s second and third levels of corporate culture, without actually acknowledging his framing.

Thomas J. Watson Sr. built a sales culture at IBM that was thriving by the end of the 1920s, and remained remarkably similar, if modified in its details, through the 1980s, resting on the Basic Beliefs. This was the theology of the firm (Schein’s second level). Behaviors, rituals, policies, and practices emanated from these beliefs in a disciplined manner worldwide (Schein’s first and third levels).

Although local national cultures influenced attitudes toward business and personal behavior, even the Dutch IBMer, Hofstede, acknowledged that a universal culture sat on top of the company as a whole.Footnote ³⁵ It was American, entrepreneurial, and laced with the importance of personal responsibility for the success of the firm. English was its universal language, much as Latin was so long ago in the Roman Catholic Church. Indeed, Catholic employees felt an affinity with IBM’s worldview almost from their earliest days in the firm. Just as a Catholic mass was the same around the world, one could walk into an IBM sales office anywhere in dozens of countries in any decade and feel immediately that they knew what people did there, how local employees worked, and how they approached issues. Like the Catholic Church, procedures were similar, ranks and titles too, as well as the rituals of “Family Dinners,” 100 Percent Clubs, Branch Manager Awards, accounting for business through “points” rather than dollars or local currency; even the way people dressed, behaved in public, and chose to live were more uniform than not. Many of these rituals (Schein’s first level activities) remained remarkably the same over many decades, particularly those that less involved customers, such as recognition of outstanding performance by employees, highly choreographed Family Dinners, and national holiday events for children. As in the Catholic Church, they used similar reports, training, data, and forms. Many communications were translated into multiple languages, from memorandum and annual reports, to anthologies of Watson Sr.’s many speeches.Footnote ³⁶

Both IBM employees and outside observers could, however, overstate the universality of such a view of the company. Students of IBM increasingly learned that regional circumstances led employees to adapt to local conditions, most often for the good, occasionally for the bad. For example, in France during the Charles de Gaulle years of the 1960s, the company was viewed as an American firm, even though usually 98 percent of its employees were French citizens, and some of the remaining 2 percent Europeans.Footnote ³⁷ IBM created a Pan-European approach to the manufacture, sales, and support of customers that presaged the arrival of a similar one by national governments with their creation of the European Union.Footnote ³⁸ IBMers were not allowed to consume alcoholic beverages at company events from 1914 to the late 1980s, yet we have one photograph of a group of Brazilian IBM salesmen at a company 100 Percent Club in the 1920s cheering the IBM photographer, each holding a mug of beer.Footnote ³⁹ By treating western Europe as one unified market, IBM was able to crush local country competitors and keep expenses lower than theirs, resulting in its national market shares often hovering between 70 and 80 percent. On the flip side of its culture, a few employees became entangled in a bribery scandal in Argentina in the 1990s contrary to IBM’s values, seen even more serious within the firm than the fact that this behavior violated U.S. and Argentinian corruption laws.Footnote ⁴⁰ How IBM managers responded to both the European common market approach and the problems in Argentina did not need to be thought out; the culture’s values dictated what their responses would be, as was the case with crises in other firms.Footnote ⁴¹

Gerstner Jr. became IBM’s CEO in the early 1990s. The board of directors brought him into the firm to resuscitate the company, then in dire straits. By Gerstner’s account, his challenge was to wrestle with IBM’s culture. Early in his tenure, Gerstner concluded that he had to alter everyone’s behavior and shift to new tactics and strategies. That meant changing the culture, because it had degraded its ability to support a vibrant enterprise conforming to his design: “Frankly, if I could have chosen not to tackle the IBM culture head-on, I probably wouldn’t have.… Perhaps the toughest nut to crack was getting IBM employees to accept that invitation [to transform]. Many used hierarchy as a crutch and were reluctant to take personal responsibility for customers.” He added, “In the end, my deepest culture-change goal was to induce employees to believe in themselves again—to believe that they had the ability to determine their own fate, and that they already knew what they needed to know.”Footnote ⁴² Successive senior leaders treaded dangerous ground if they mishandled IBM’s culture in the opinion of employees.Footnote ⁴³

Second, IBM as an Information Ecosystem

In his studies of managerial practices in the 1960s and 1970s, Chandler repeatedly reported on the growing use of information—largely operational and accounting data—that made it possible to control and coordinate the work of large enterprises, such as railroads. Naomi R. Lamoreaux and Daniel M. G. Raff focused additional attention on the role of information in their study of coordination and information practices in the mid-1980s.Footnote ⁴⁴ Yates described links between such information and communications as both operational and technological, which made possible the collection, use, and storage of relevant information. Heide, Cortada, and others added descriptions of how the use of tabulating and computing equipment permeated corporations across entire industries and countries.Footnote ⁴⁵ Economists and some historians began speaking of the diffusion of computing around the world, increasingly documenting the massive and rapid appropriation of these information and communications technologies that became, in effect, hardwired, permanent features of organizations, analogous to nervous systems in live creatures that, too, transported information all over the body.Footnote ⁴⁶ The majority of their focus concentrated on describing specific types of data.Footnote ⁴⁷ Historians examined engineering and scientific data and sales information, and most recently Yates, again, waded into the discussion with Craig N. Murphy by examining technical standard setting since the mid-nineteenth century.Footnote ⁴⁸ In short, historians have long looked at specific kinds of information exchange, whereas others explored the nature of corporate cultures, which are entwined with information.Footnote ⁴⁹

Their interest raises the following questions: Might they benefit from combining both lines of investigation more overtly, thinking more broadly about the role of information? Can viewing a company’s activities through the lens of an information ecosystem advance the work done over the past half century on the role of data, communications, and managerial practices of the firm? IBM’s case suggests yes. Accepting the evidence that IBM had a highly articulated corporate culture with a set of well-thought-out values, processes, and long standing traditions begs the following question: How did that happen? The emerging new subdiscipline of information history is beginning to provide answers. Collections of information from within and outside the enterprise, from customers, employees, advisors, markets, competitors, associations, regulators, and so forth allows us to track its emergence. IBM’s information ecosystems existed within the enterprise and external to it, which supported the collection, distribution, and application of information.Footnote ⁵⁰

In the case of IBM, the enterprise had to define what information it wanted to collect and use (e.g., financial and revenue, sales too); make its application routine practice (e.g., monthly, quarterly reporting); and determine its influence on the firm’s behavior (e.g., how it sold, designed, and built equipment and software, what strategies it embraced, how it made decisions). Those elements required strategy, creation and sale of products, a corporate culture aligned with IBM’s strategy, and then the experiences of successes and failures buttressed by the use of hard data (e.g., numbers, statistics). Economist Douglass C. North explained the interaction, for example, between culture and knowledge (i.e., information): “Culture provides a language-based conceptual framework for encoding and interpreting the information that the senses are presenting to the brain” and “processing information is the key to understanding a more complex behavioral pattern.”Footnote ⁵¹ IBM’s management did this, defining and integrating the core pieces of these at all levels of the firm and through its operating practices, beginning in the 1910s and essentially completing its deployment by the start of World War II.

But no company operated in isolation, so where does IBM’s information ecosystem begin and end? In answering this question it would be useful, indeed practical and realistic, to include the information and communications of customers, competitors, regulators, law courts, standards setting bodies, industries, industry associations, media, and families of employees and customers, because employees interacted with these various constituencies, sharing information back and forth in constant communications. For example, in examining IBM’s antitrust litigation experiences from the 1930s to the 1980s—mapping out what information it collected and used, how it was transferred to law firms and government prosecutors, and how it was circulated within the firm—signals to the historian that the information needed to navigate litigation was far more than normal legal considerations. It involved customers, all parts of IBM, the media, and entire communities (e.g., Poughkeepsie, New York, where mainframes were made). In such a worldview, lawyers have to fight to stay in the center of the narrative. That approach helps explain why, for example, merely announcing that millions of pages of documents were submitted in court cases did not begin to describe the volume and breadth of information that so many individuals relied upon.

Beginning inside IBM, then expanding out to describe its links to customers and suppliers, and then further out to include industries, rivals, and associations, allows an historian to describe the company’s world. This is an iterative process, as no single historian can do it all. However, it begins by recognizing that when one studies information ecosystems, the simplicity of describing, say, the evolution of a company’s accounting practices, such as has been done so for DuPont, for example, proves inadequate to the task; it leaves out too much. The IBM case demonstrates far more information affected any category of other information, such as sales metrics or financial reports. The company recognized that reality as early as the 1910s when its new general manager, Watson Sr., implemented quickly at C-T-R (future IBM) an array of data collection procedures he had used in his prior position as senior sales executive at NCR. His “dashboard” (to use a modern term) allowed him to see a combination of information about sales revenues, costs of achieving these, the number and types of meetings his salesmen had with customers on a weekly basis, and other facts about how sales regions operated. Sales managers groused that he knew more about their territories than they did. Without using the phrase “information ecosystem,” with its implied inclusion of many elements, Watson had created one. It proved so successful that he and others added more data elements to it over the decades.

Business and technology historians continue to shift from looking at just the behavior of entire divisions (or specific individuals) to what information exists, how it is moved within an enterprise, and how it is used.Footnote ⁵² Elsewhere I have started to describe IBM’s information ecosystem, but briefly summarized, that description includes identifying what data was collected by each division, department, or branch office, what reports were routinely prepared, how these data and reports were used, and their effects on the performance of individuals and groups within the firm.Footnote ⁵³ That exercise builds on the early work of Yates and others, applying it to a specific company. Through this lens, we can understand the role of accounting data and financial balance sheets, comprehend how employees wrote annual reports and made other filings with governments, and appreciate major protagonists and newsletters as vehicles by which many managerial messages were delivered to employees, and so forth. Exploring the role of entire industries that way, however, remains largely unexplored terrain for business historians.Footnote ⁵⁴

The objects themselves become less important than the content they hold. What people do with their training, results, and experiences entwine to display the activities of the company. Elsewhere, for example, I explored what information salesmen had and used from the 1920s through the 1980s: product descriptions; company newsletters; data on their sales performance; required training (for transmitting to them information of specific types); performance reviews; roundtable discussions; interactions with customers; and participation in industry events. Their information ecosystems and sources of facts were complex, thick, and valuable.Footnote ⁵⁵ Case studies of IBM’s internal information infrastructures and how they evolved are redefining what we understood occurred within the firm.Footnote ⁵⁶

Viewing IBM’s information ecosystem led to an exploration of whether the firm’s management understood that reality and purposefully leveraged it for competitive advantage. The answer was a resounding “yes” in all periods. Thomas J. Watson Sr. sought from his earliest days in the firm to shape his customers’ perspective about data processing, using publications, presentations, and speeches by thousands of employees. He and employees across the firm trained customers on why and how to use data and his machines to control and expand their own information ecosystems, which served to benefit IBM as well.Footnote ⁵⁷ Successive generations of employees did the same, normally staffing industry committees and associations, maintaining a lobbying operation first in Washington, D.C., and later in other capital cities beginning in the 1930s, and having its engineers and computer scientists publishing extensively about computing technology.Footnote ⁵⁸ In IBM’s case, the ecological lens offered a substantially new perspective on how the firm operated that differed from (yet complimented) the work of prior historians looking at the company.Footnote ⁵⁹ It became possible to link Watson’s thinking over the course of a half century to Usselman’s discussions about IBM’s intentions in the 1960s and 1970s, to Gerstner’s of the 1990s, to the technological studies conducted by Emerson Pugh and his colleagues about the evolution of IBM’s mainframes to that of PCs, and more recently, to IBM’s marketing efforts through its traditional marketing and advertising initiatives, simultaneous with the research done by the IBM Institute for Business Value, its managerial think tank. The role of IBM product user communities now fit better into our understanding of IBM’s effect on customers, such as those involved with SHARE and other associations.Footnote ⁶⁰

Walter Friedman described how, as sellers in American companies professionalized, they increasingly collected information in structured ways that they communicated into their firms to be used for the development of products, to understand and counter competitors, and to shape terms and conditions of contracts and pricing.Footnote ⁶¹ A quick takeaway from reading his history is that this same professionalizing activity of harvesting information was not limited to sellers. In fact, in his next book, Friedman documented how corporations purchased whole bodies of accounting, financial, and other economic information.Footnote ⁶² In studying a company like IBM, one could ask who else collected and used information on a regular basis from outside the firm. In IBM’s case, almost every department and division did. One conclusion that could be reached is that the firm proved effective for so many decades in part because so many employees “listened” to anyone affecting the company. Usselman made that point: “IBM was an organization whose business had naturally fostered.… salesmanship that required close attention both to technology and to the particular requirements of each customer, regular exchange of information between the field and the plant, flexibility in production, and a willingness to compromise.” Usselman attributed such behavior to IBM’s success in entering the computer business in the 1950s and succeeding in it in the 1960s.Footnote ⁶³

By mapping the information available to employees and tracking how it was brought into or exported out of the firm and by whom, it becomes clearer that in all periods, management understood a great deal about their business realities, how technologies were evolving, and their competitors. In other words, successes and failures in the crafting of strategies could not be solely attributed to misunderstanding their markets. Other explanations were required. For example, their inability to accept that their markets were shifting from just mainframe centralized computing to PC-based distributed processing in the 1980s that led to the near collapse of the firm in the early 1990s cannot be blamed on lack of information. The disaster has to be understood through other means, in this case by understanding the influence senior executives were subjected to by their heritage of growing up with mainframes, which led individual leaders to deny the emerging market realities their staffs described. It is that kind of insight that links together corporate culture, sociological approaches advocated by post-Chandlerian historians, and information ecosystems.

An X-ray of IBM’s Information Ecosystem

This lens takes on some features of an x-ray of the firm. A classic organization chart is an example. Figure 1 is a simple illustration of an informational x-ray. Viewing IBM through an x-ray-like approach makes it possible to do the same for other relevant organizations with which it interacted, such as customers and industry associations, each with its own graphical representations. These x-rays can be buttressed with content by hunting for what information resided in each, discussing the nature of their reports and other documentary remains, and working out how people used these materials.

Figure 1

How might one describe IBM through such a lens? This figure illustrates several types of information. First, each of the boxes represents communities within and outside of IBM that create, use, and are influenced in their behavior by information they received. These boxes represent significant communities within IBM’s information ecosystem. The internal ones were actual entities with organization charts and documented chains of command. Second, arrows point in the direction that employees and others sent most information and where this information had the greatest influence, recognizing, however, that information moved back and forth in each case, and sometimes between other entities within this ecosystem to a lesser extent. These arrows and their direction were determined based on evidentiary observation, not assumptions.Footnote ⁶⁴ Third, one could, by extension, build out more complete, evermore complex detailed networks, such as from industry associations to other vendors and to customers. For our purposes, this graphic is IBM centered.Footnote ⁶⁵

Since at least the mid-1950s, employees were in the habit of creating organization charts that they almost universally included in their formal presentations to one another, to customers, and to industry associations. These always described where one’s own personal organization fit in the greater IBM, such as a sales manager’s chart showing his or her sales unit within a branch office’s organization and then within the branch office’s position in a sales district, and even within the sales division in a specific country organization. The word “universal” is intended because, in examining presentations or participating in thousands of meetings, almost every slide deck included such organization charts. In fact, when Gerstner came to IBM, he complained about the inability of his executives to speak without using charts, all of which included such material.Footnote ⁶⁶ Salesmen and technologists working with customers presented similar information about these external groups, in some instances, creating charts that even their customers found informative about their own enterprises, such as how a particular data processing department was organized.Footnote ⁶⁷

Fourth, if Figure 1 suggests that the movement of information was static, that is intentional, because as the dates listed in the figure suggest, each participant and every suggested direction of the distribution of information remained essentially the same from the mid- to late 1920s through the 1980s. What changed were specific pieces of information (e.g., about computers instead of tabulators, formats for presenting forecasts); the frequency with which these were transmitted back and forth (e.g., tended to increase in frequency from, say, sales forecasts quarterly in the 1920s to monthly by the 1940s, to frequently weekly by the end of the 1980s); and the tools used to do that (e.g., mail in the 1920s, telegraph as well, IBM 2741 terminals between the 1950s and mid-1970s, “dumb” “green screen” terminals with telephone lines in the 1970s and 1980s, PCs and intranets in the 1990s). Over time, new types of information were added that went to different places not shown on this graphic, such as travel expense reimbursement forms from employees to their office’s administrative department (1920s–1970s), then, in the United States, from all sales personnel to a dedicated reimbursement center in Endicott, New York.Footnote ⁶⁸ These documents were paper based and hardly changed from the 1920s through the 1970s, and then were submitted increasingly online, capturing the same classes of data as in earlier times, often in the same format.Footnote ⁶⁹

One could identify the obsession with information across the enterprise. For example, look at the work of repair personnel, called variously field or customer engineers. From the earliest days of the company these were the individuals who installed and repaired hardware, IBM’s equivalent of automobile mechanics, both in how they worked and in their attitude toward “paperwork.” They excelled at the first and found the second irritating and intrusive of their time. Nevertheless, they and the company lived by their data collection. For decades they were required to fill out an eighty-column card called an “Incident Report” (IR) every time they touched a machine, and it had to be returned to the company that very day. IRs included the name of the customer, customer identity number, name of the engineer and his employee serial number, amount of time spent doing the work, when, and type of activity. If it was billable work then they added time, expenses incurred, part numbers, and so forth. For their activities, they added nearly a dozen additional pieces of information (e.g., symptom of a problem, names of parts involved, time spent on these, and location of the work). On this card one saw clusters of information gathered about billable activities, the activities themselves, identification and travel of participants, parts, and which organizations within IBM were involved (e.g., branch office of the customer engineer, plant number).Footnote ⁷⁰ There was a similar card if installing a machine, on which a field engineer would explain what he had installed, where and how long it took, and so forth; most important for accountants and sales personnel, he coded that the machine was now functioning properly and had been turned over to the customer. That meant a customer could now be billed for rental or for its purchase. To trigger that financial activity, the customer engineer had to “code 20” the machine, that is to say, pick from among his options on status the number 20. For generations, salesmen wanted to know when a machine had been “code 20ed.”

A card reader trapped that information, which then was collated and processed, integrated into reports that routinely went to sales and field engineering offices, and summarized and moved up the organization and over to the manufacturing plants that made and supplied the parts and machines. These cards were analyzed much the way today’s technologists analyze big data, statistically and analytically looking for patterns of behavior and usage in order to forecast demand for machines, manpower, parts, and expenses. In the 1980s, portable communicating devices called “bricks” because of their shape, which presaged the tablets of the early 2000s, replaced cards. Information about the features and maintenance history of individual machines was collected into a system called “Retain,” used by manufacturing plants and design engineers to design new machines, improve existing ones, or shape future manufacturing and repair practices. Reviews of the kind of information IRs revealed were also presented in forms similar to Figure 1, almost as if the data were organizations and people themselves.

Does this graphic represent a generic one evident in other companies? I believe not.Footnote ⁷¹ Each company operates in an information ecosystem unique unto itself, made so by who worked there and by what information they relied upon. One could make the same argument with respect to what such a graphical representation would show, say, of IBM’s manufacturing community. On such a graphic, a manufacturing site would be in the middle of the illustration and the sellers on the fringe along with customer staffs and branch offices.Footnote ⁷²

Figure 2 suggests that if one thinks of IBM as the center of attention in its information ecosystem, we can see that information moved out of the company in waves, initially and immediately to those of direct concern to the firm, such as customers, competitors, suppliers, and industry associations. These represent communities that the largest numbers of employees encountered most frequently—daily being normal, hourly in some instances. The next ring represents those communities that IBM interacted with less frequently but were still relevant to its operations. This chart, more than Figure 1, is quite generic in that it could serve as a template of what happened in other firms, such as at Ford Motor Company, Nissan Motors, Sears, and AT&T, all enterprises with which this author conducted consulting projects. Again, it is a tool that assists in defining the information ecosystem of a company.

Figure 2

The concept of waves is so useful that, at the risk of some throat clearing, we should pause to appreciate it. The concept has a long history among economists, but they thought of it as waves crashing onto beaches from oceans. Here, I mean something more akin to when one throws a stone into a lake. This is an important distinction because, in the lake version, we are identifying a source from which information is going out. Information is purposefully and usually consistently sent across an organization (i.e., IBM) out to customers, industry associations, possibly society (e.g., as happened with IBM’s PC and how people should use these in the 1980s). Second, the graphic means to suggest that communications to the greater world are centralized, and in IBM’s case and that of other multinationals, it was and continues to be the case. It is also a salute to the ideas offered up by Lamareaux and Raff concerning “coordination.”Footnote ⁷³ Their notions apply well when studying corporate information ecosystems as more than putting out marketing, advertising, or operational guidelines.

At IBM, the notion of waves draws our attention to patterns of information sharing tied, for example, to calendars. Beginning in the 1910s and extending to the present, all sales offices and their various levels of organization, up to the CEO, regularly reported on their performance, usually monthly, with annual, more extensive multiday reviews of past, present, and intended performances by account, sales territory, geography, and country. Product developers and manufacturing had their own cadence of meetings and seasons, all synchronized one to the other to create more composite views of the company’s activities, the higher the presentations were made. These routinely looked back at prior performance and addressed issues of what was coming next—hence the notion of waves—and had a relatively standard format applied year after year. That facilitated comparisons of performance and provided a structure for understanding many disparate pieces of information. The same technique was used with customers and industries. For example, IBM speakers made presentations at technical and industry conferences about “trends and directions” in technology, a code name for what IBM was working on and would introduce in the future as new products without actually committing to doing so until the timing was right.Footnote ⁷⁴ Making such presentations at conferences had become routine by the mid-1920s, pervasively with customers in the same period. It did not happen much earlier because Herman Hollerith, in charge of tabulating sales before Watson’s arrival, displayed little interest in briefing customers on a regular basis, preferring to discuss his products once a sales opportunity had surfaced.Footnote ⁷⁵

Now turn to some of the content symbolized in Figure 1. Between 1911 and 1950, IBM conducted the majority of its product development in Endicott, New York, where engineers designed new machines, often suggested by salesmen and their customers, both of whom routinely visited the manufacturing site located within walking distance of R&D. This feedback was continuous across the nearly half century of the tabulating equipment business and continued as a practice at all plants and laboratories into the twenty-first century.Footnote ⁷⁶ So, the content and behaviors reflected in this figure were not limited to sales.

A second locus of information handling existed at corporate headquarters in New York City, where staff collected information about what competitors were selling and at what prices, data they fed to Endicott and other plant sites to help shape product innovations and pricing. Corporate personnel collected financial and sales data from around the world, country by country, which they used to inform management on how much in dividends to pay out, allocate budgets for all departments, and shape the business strategy for the firm.Footnote ⁷⁷ For most of IBM’s history, CEOs operated a senior management committee, whose name changed periodically, but that met usually one or more times per week to adjudicate distribution of budgets and personnel across competing divisions and geographies, establish company-wide operational practices, and approve development and introduction of new products. In periods of intense change, these bodies of information and senior management meetings proved crucial. It was this confluence of information and executives that led to the decisions in the 1950s to move into the computer business, more as a collective process than simply as the heroic initiative of one or few executives, as tradition held. The same occurred again in the early 1960s with the seminal decisions to build the System 360 family of computers, the product that rocketed IBM into its golden age of mainframe sales. Conversely, but only slightly less so, that same collective of information, and managerial use of it, shaped the initial decision and approach to the development of the IBM PC at the dawn of the 1980s. Waves of information crashing at the doorsteps of these management committees were unending, literally every week for over a century.

Perhaps the most visible example of the information ecosystem at work was the collaborate efforts of IBM’s board of directors and corporate headquarters in the publication of their annual reports since 1911, and after World War II of such subsidiary documents as the 8-K and 10-K reports required by the American government. Between 1911 and the late 1940s, annual reports were simply short financial statements covering such items as revenues, gross and net profits, and expenses. Toward the end of the 1940s, large corporations changed their annual reports to accommodate letters from the chairmen explaining the firm’s business, its strategies, and its role in society; IBM did the same. That is when it added lengthy articles showing off its new products, photographs of customers posing next to computers and later PCs, and updates for readers on evolving business conditions. IBM distributed these reports by the millions to stockholders, media, academics, employees, customers, and anyone else who cared to see them. Their text aligned with the messages delivered by executives, marketing, and advertising departments. Analysts and employees learned to read these carefully, especially the chairman’s letter, for indications of what might be done next, particularly regarding strategy. These documents contained some of the most important sets of statistics and statements of intentions that IBM delivered to the world, and specifically to its employees and customers, becoming annual additions to IBM’s information ecosystem. The basic data would be announced in late January, and people would be able to read the full reports in April. When communicating with customers, pithy quotes from the chairman’s most recent letter invariably ended up on slide decks used by IBM employees. By the end of the century, it was not uncommon for these reports to be over one hundred pages in length, and later even longer, as regulatory requirements increased for the disclosure of financial information.

For the historian, these reports became increasingly revelatory regarding both financial performance and understanding, at a macro level, corporate strategy across the past century, especially when it became apparent that some IBM country organizations published local versions as well, such as IBM France about French operations, beginning no later than the early 1960s.Footnote ⁷⁸ Whereas business historians have long paid attention to annual reports, these should not be viewed as one-offs from corporate or a board of directors, but as part of an integrated component of an information ecosystem. Seen that way, their role within the firm becomes clearer, while their increased marketing role with customers does too.

However, up and down the organization, employees gathered similar information. Each IBM country organization focused more on what individual customer enterprises needed and used from IBM, collecting this data on a weekly and monthly basis to inform actions of the country management team. At the branch office level, salesmen kept files on each account, containing information useful to their sales operations. These included biographies of customers, inventories of installed IBM products, information on how they were used, accounts receivable records, and customer publications, such as their annual reports and newsletters.Footnote ⁷⁹

These bodies of information were used to marshal resources to the advantage of particular constituencies inside the firm, usually divisions, sometimes headquarter departments. IBM salesmen used such information to identify opportunities to use IBM products in their accounts; to develop documented plans for what they were going to sell and what they were going to teach their customers; to prepare forecasts for business they intended to sell (“close”) that month and quarter; to explain what resources they needed to carry out their work; to keep track of equipment on order; to monitor accounts receivables; to document understandings with customers via exchanges of letters; and to maintain a file of signed contracts, customer satisfaction survey data, and all proposals. They used these various collections of information to craft plans for what to sell to customers; to inform them on how to support their activities; to request assistance from other parts of IBM; to explain to IBM management what they were doing with customers and how those activities were resulting in new business; and to respond to requests for information regarding customer needs, forecasts, and insights about the industry or company.

Because this body of material reflected a complex information ecosystem, salesmen and their chain of command spent at least a third of their time shuffling information in and out of the corporation and up and down their organizations. Often customers, too, wanted their IBM representatives to provide similar information to them, in addition to briefings on “trends and directions” in technology and IBM’s operations.Footnote ⁸⁰ It was not uncommon for IBMers to explain to their customers what the latter had acquired, what IBM intended to sell to them next year, and why.

Although graphical representations look two dimensional, in reality they are more complex. With respect to Figure 2, the number of places where these general activities occurred increased, from several tiny plants to over a dozen research and manufacturing facilities, from a couple of dozen sales offices to over 800 worldwide, from a handful of countries to over 150 by the end of the 1980s.Footnote ⁸¹ Not depicted here were the growing interactions among peer organizations within IBM’s ecosystem: R&D sites talking to one another and coordinating worldwide development of products (beginning in the 1930s); sales offices coordinating operations with multinational customers that expected IBM to collaborate (i.e., coordinate) with all their various data centers (beginning in the 1950s); and manufacturing facilities responsible worldwide for product support, such as the software missions of Hursley, England, and Poughkeepsie, New York, for mainframe software operating systems (largely starting in the early 1960s).Footnote ⁸² Thus, a more accurate graphical representation of the movement of information would have to be holographic, more than even three-dimensional.

Historians more comfortably refer to such movements of information as communications. However, it may be more accurate to characterize these activities as the shuttling of data back and forth through ebbs and flows of anthropomorphic facts (that is to say, information intelligible to humans) while sensors increasingly also moved electrical impulses from one place to another. Today, AI increasingly is authorized to react to and to make decisions much as humans did in the twentieth century. ATMs, for example, did this and made decisions whether or not one had enough funds in their bank account to permit humans to withdraw money from the bank. Information that flowed through an enterprise was ritualized with common “look-and-feel” formats standardized across companies; IBM was no exception over the course of many years. That combination of standardized format and types of information routinely delivered in some specified manner made it possible for these to be consolidated by fewer people, increasingly through automation. Those actions made it possible for the ecosystems’ intelligence to work its way up and across IBM. Recall the standard formats for a salesman’s business forecast or a field engineer’s IR cards. For decades, even annual reports had a similar structure of anticipated chapters and sections, even though the data changed each year along with the photographs and feature articles.

Information had its seasons, too. Recall that for decades the entire company from bottom to top annually put together “spring” and “fall” plans in which organizations articulated their business plans, results to date, anticipation of what would happen in the remainder of that year and next, resources required, or problems needing resolution.Footnote ⁸³ IBM used its own products from its earliest days to facilitate collection and use of such information, largely numeric in the early decades, but by the 1970s textual and graphical as well. Anything that could be measured seemed to be.Footnote ⁸⁴

Implications

Looking at IBM as an information ecosystem suggests several implications for those studying the history of any firm, not just long-lived ones or those, say, in a “high-tech” industry. Using an information lens worked effectively in enriching a more sociological/anthropological approach to the study of IBM’s history. There is a convincing logic to that, because IBM and other firms spill over into so many types of activities, themes, and questions that it is nearly impossible to avoid discussing together revenues, technologies, leaders, communities, customers, regulators, society, information flows, and so forth. However, the approach suggested in this essay can provide both business historians and students of a company disciplined approaches—think checklists—of topics that should be looked for and understood—a tip of the hat to Schein, Yates, Lamareaux and Raff, and increasingly to those business historians now publishing, for example, in Information & Culture, who favor more precise taxonomies bounded by chronology. There are minimally three reasons for historians to view corporations, and by extension industries, through the lens of information ecosystems. These can be explained by the new light shed on IBM’s history by its use.

First, looking at what information was collected and how it was used expanded our understanding of how integrated were the various parts of IBM. The amount of coordination across vast organizations and geographies became stunningly clearer. No longer did we need to rely on glib executive sound bites, rather, we now had a level of specificity that gave us a greater understanding of how other large enterprises potentially operated, particularly regarding the intimacy between product development and sales. It was common in the 1970s and 1980s for employees, in surveys regarding how they spent their time, to speak about “paperwork” and other “administrative” activities. They were probably correct in estimating how much time they spent on these matters (usually a third or more), but the approach for studying the company presented in this essay told us on what they spent this time and, perhaps more important, why, which was so important in revealing how such a large enterprise was able to function profitably.

Second, historians had known superficially that IBM sales tended to be a rather conservative branch of the firm, despite its public image of being the “finest” and the most “successful,” “clever,” and so forth. Chasing their paper trail highlighted their resistance to change if it threatened “account control” over customers. Additionally, IBM sales were not as successful as we were often led to believe, operating as heroic lone warriors; rather, they succeeded because so many other parts of IBM did too. These combined successes included detailed, ongoing exchanges of ideas between field and factory, the integration of financial and product pricing strategies with appropriations of new technologies done in a timely fashion. IBM’s success was always a team sport, and the information trail demonstrated how true that was. These teaming activities documented the many statements made by executives and the most senior leaders about the value of teams and collaboration, forcing us to realize that these were not platitudes, but rather serious statements about how things really worked inside the firm. As a corollary, when some executive went rogue on collaboration, or did not share information, one or a few individuals could upset what was a delicate collection of operational practices. It is why, for example, CEO Frank Cary needed to, and was able to, isolate the development of IBM’s PC out of the normal product development process.

Third, customers were more tied to IBM’s way of thinking and working than we knew before. Prior, historians and industry pundits spoke of their dependence, more about their need to use IBM equipment because it was too expensive to switch to another vendor’s—the path dependency argument—or because the hardware and software did the job they needed done—the “nobody ever got fired for using IBM equipment” argument. However, the information ecosystem lens taught us far more. It showed us that the dialogue about ideas was a two-way conversation back to the late 1890s, between customers and employees who were both equally versed in business and technological issues and not shy about expressing their needs and other points of view. There probably was never an IBM salesman or field engineer who did not hear from a customer about what IBM should be doing next and why. The industry conferences, user groups (e.g., SHARE), and those visits (“calls”) every executive was required to make on customers since the dawn of the company’s founding all spoke to the importance of that back-and-forth dialogue. Before, we simply did not know how extensive that was, nor much about what they discussed. Customers wanted IBM to succeed and they often explained to IBM how to do that. When the president of the Metropolitan Life Insurance Company told Tom Watson Jr. that he was filling up warehouses with cards and needed to move to magnetic tape storage in the 1950s, he was telling IBM what to fix so that the two of them could use computing more effectively.Footnote ⁸⁵ When IBM’s role in shaping the world’s view of data processing came up in history books or in the media, it was usually about IBM telling a customer what to do, how to think about an issue. The argument went that such “thought leadership” was what made IBM an iconic global corporation. The information lens show us that it was not just IBM lecturing customers so much as listening (Usselman’s point too), sharing insights about technology and its uses that kept constantly unfolding into new forms. That is why understanding the informational world surrounding IBM is so important to appreciate. Stopping at the fence line of IBM’s corporate boundaries leaves out too much of the story of this company. The findings regarding IBM are applicable to other large and midsize enterprises.