The Real Problem with Computers

The Squandered Computer: Evaluating the Business Alignment of Information Technologies Paul A. Strassmann New Canaan, Connecticut: The Information Economics Press, 1997 Information Ecology: Mastering the Information and Knowledge Environment Thomas H. Davenport with Laurence Prusak New York: Oxford University Press, 1997 Businesses worldwide—and particularly in the United States—have wasted billions of dollars believing the big lie of the Information Age. For almost two decades, that lie has encouraged a massive spending binge, absorbing over half of every dollar that U.S. business has invested in itself. And it has governed the ways in which both entrepreneurs and global juggernauts seek competitive advantage. The big lie is pervasive, and it offers a seductive logic that actually makes it believable. The lie says that if organizations only had greater quantities of cheaper, faster, and more useful information, they could increase their profitability and enhance their competitive positions in the global marketplace. On the surface, that makes sense. If you offer employees greater quantities of better information more quickly and at a lower cost, you should reasonably expect their performance to improve as a result. Although in many situations better performance will result, the sad reality is that even improved information often has little or no impact on people’s behavior. Who is unaware of the risks of smoking? Yet millions of people still pick up the habit. Yes, there should be strong links between information and behavior in the enterprise. But let’s be honest: the real problem most executives face isn’t inadequate information—it’s the organization’s unwillingness to change behavior in the face of good information. This failure on the part of U.S. business brings to mind the comment made by English biologist T.H. Huxley that there is no sadder sight than an ugly fact slaying a beautiful hypothesis. Two recent books on information management take on and refute the “beautiful hypothesis” about the business value of information technology. Both conclude that managers have acted irresponsibly in relying on technology to solve fundamental problems. While Paul Strassmann’s Squandered Computer urges executives and technologists to confront bracing economic calculations about enterprise computation, Thomas Davenport’s Information Ecology offers the framework for a solution. Perfect complements, these books neatly capture and present the disturbing information and technology issues that most organizations would rather not discuss. Grim Calculations Strassmann, formerly the chief information officer at both Xerox and the Pentagon, presents his analysis with the zeal of a convert. He challenges the idea that U.S. corporate investment in computing has generated appropriate returns and flays “best-practice” information technology departments as shams. He tests the core of the arguments driving IT, only to hear the echoes of hollow claims. When he scrapes away the muck from the accepted methodologies and trims some of the statistical assumptions, he finds that executives’ faith in information technology has cost U.S. business tens of billions of wasted dollars. Not only does Strassmann find no demonstrable relationship between spending on computers and corporate profits, but he also rejects recent findings that computers have finally begun to boost the economy’s productivity. He argues that the recent rise in revenue per employee is more likely a reflection of outsourcing than of gains attributable to computerization. Anyone who has gone through Economics 101 and a basic statistics course will find Strassmann’s math a provocative and welcome antidote to popular media coverage, which regularly attributes 1,000 % returns to intranet investments. On a micro level—even on an industry-wide level—it is inarguable that some companies get strong returns on their digital investments. What seems true, however, is that on a macro level more money has been wasted on computerization than has been created by it. On a macro level, more money may have been wasted on computerization than created by it. No one denies—including Strassmann—that computerization and creative networking can add enormous value. But when we look at the numbers, it is clear that companies are not basing their computer investments on careful calculations of returns or added value. Other factors such as culture, politics, fashion, and competition also come into play. Strassmann persuasively argues that best-practice methodologies often are irrelevant benchmarks for many companies investing tens or hundreds of millions of dollars in computers and networks. As for the popular trend of IT out-sourcing, Strassmann argues that it is more an instrument for downsizing than a process for adding value. He has found that companies with consistently large profits and rising employment have not outsourced most of their information technologies, despite the supposed benefits of synergy or the advantages of getting rid of commodity work. The only companies that have outsourced information technologies are those that are shrinking anyway. For every success story—for every Federal Express and SABRE—just how many IT failures have arisen from popular delusions about computers? For every large-scale project that has been implemented successfully, how many have been canceled or have ended inconclusively? The answers are chilling, as Strassmann shows in his extrapolations from survey data about failed software projects. Indeed, employees can misuse computers in a number of ways. How many times can people revise a word-processing document, manipulate a spreadsheet, or polish a presentation before they hit the point of diminishing returns on their time? Is productivity really increased in this context? When you multiply all those diminishing returns by the tens of thousands of personal computers in a large company, what is the real value-to-waste ratio? Strassmann doesn’t treat these questions as hypothetical. Precisely because his perspective on the diffusion of technology is more statistical than anecdotal, his numbers have a credibility that the Information Age prophets frequently lack. Moreover, it would be wrong to dismiss Strassmann as a cynic or a corporate neo-Luddite. On the contrary, he comes off as a wounded idealist. He appreciates the potential of information technology but simply refuses to ignore what happens when managers misuse that potential. Strassmann’s implicit message is that the profitable use of technology does not begin with a better understanding of digital media; it comes with a better understanding of the organizations that use them. Systems Are Not Enough Where Strassmann offers persuasive numbers, Davenport portrays how the use of information depends on an organization’s imperatives. He begins with the overused cliché that “technology is not enough”—that people are the key to effective information creation and management. But the body of the book uses anecdotes and case studies to reveal the organizational, political, and behavioral elements that influence how enterprises process information. More than just a metaphor, the term information ecology represents Davenport’s genuine sensibility about how companies operate. There’s a fundamental difference between managing an information system and running an information ecology, just as there’s a difference between operating a grape press and making wine. Managers intent on establishing technical systems subscribe to different values and practices than managers trying to set up productive environments for their workers. Ecologies of information have a much broader array of interaction and interdependence than systems of information. When managers try to fit inflexible, mechanistic systems into organic contexts, they need new vocabularies to explain how people in organizations really use these systems. Indeed, the word information loses its edge and luster when redefined in ecological contexts; culture and politics—and relationships generally—become at least as important. There’s a fundamental difference between managing an information system and running an information ecology, just as there’s a difference between operating a grape press and making wine. Davenport demonstrates this awkward phenomenon in a hypothetical example of how centralized databases get in the way of employees’ creativity: Trying to improve sales, a manager begins to look into her company’s distribution process and, in doing so, decides to add distributors to the company’s list of customers. Because her expanded definition of the term customer conflicts with the established usage in the customer database, the manager creates her own database, which she then shares with others. Over time, multiple labels for a primary unit of information proliferate and undermine the utility of the common database. But Davenport says such muddying of the waters can be useful: “One might even argue that if a firm’s employees are not attempting to proliferate meanings for key information entities, they don’t really care.” He concludes that “the more an organization knows about a term or concept relevant to its business, the less likely it is to agree on a common term or meaning for it.” That observation is both provocative and profound. Complex organizations construct and manage meaning in intricate ways. Precisely because the same word means different things to different people, the quality of the organization’s relationships, more than the quality of its information, will determine how problems can be solved or opportunities exploited. Davenport’s observations show that human relationships create a context for information more than information provides a context for human relationships. Nevertheless, much of Davenport’s advice consists of telling managers how they can change their organization’s culture to fit the technology. His focus on information can be a bit like a cookbook that emphasizes the role of pots and oven settings. Surely they matter. But what about the food? What about the taste? Ecologies are more about relationships and interaction than about information. The Freedom to Focus on Relationships Consider this conceptual experiment as a way to rethink information ecology: Every time you see the word information in a book or article about technology, substitute the word relationship. Notice how that simple substitution radically changes the design sensibility underlying intranets and other management-information systems. When we design networks for the enterprise, are we designing them to manage important information or to manage important relationships? The answer, of course, is both, but our anachronistic Information Age vocabularies distort what is really going on. Though Davenport’s interpretations and insights revolve more around the contexts of human relationships than the content of information, his book still presents an infocentric view of the enterprise. The real struggle is not over information, but over relationships. Once we understand the importance of relationships, we can start to use technology differently within the enterprise. A decade ago, companies making strategic investments in information technology had only a few circumscribed choices from which to work. What the organization wanted was secondary to what the technology could do. Tomorrow’s predicament is the opposite: Organizations will have more choices and opportunities than they know what to do with. They will have to decide what they really want to do. Technology will no longer be the main obstacle; the range of choices will be. As Tom Malone, professor at the Massachusetts Institute of Technology’s Sloan School of Management, observes, “The more choices an organization has, the more its values matter.” Does the organization want to use its networks to centralize or decentralize responsibility? Does the enterprise want to make every bit of data accessible to everyone all the time? Or does it want to build a new information-access hierarchy into its intranet? Should individuals be rewarded for sharing information? Should people be encouraged to strike up electronic relationships with employees in other departments? Or should interdepartmental fraternization be deemed an inappropriate use of the network? Those are serious questions. As years go by, technical issues will continue to yield dramatically to business, cultural, and organizational concerns. It should come as no surprise that “peopleware” tensions—not hardware and software disputes—will dominate the design discussions that truly matter. In other words, the economics of enterprise computing have been turned inside out. Five years ago, organizations struggled with scarcities of processing power, bandwidth, and distribution. Within another five years, those same organizations will face an embarrassment of riches. The future of digital technologies will be one of more choice for less money. Even today, advanced enterprises find that the marginal costs of network expansion are diminishing. Ironically—and appropriately—organizations are discovering that the value they gain from their technology investment depends on the choices they make. As Davenport rightly observes, collaborative computing does not create collaborative cultures. Collaborative tools must be aligned with collaborative incentives—a pervasive theme in the book. Why should someone post useful information on a groupware database if someone else can expropriate that data for his or her own purposes? If two divisions are maneuvering for the same slice of the corporate budget, just how accessible do they want to make their budget proposals to their in-house rivals? Do the benefits of internal cooperation outweigh the costs of competition? If an organization does decide to improve the way it shares information, it should focus first on changing the culture of sharing. Most information managers know little about designing incentives for enterprise collaboration. That’s why responsible information departments have to insist from the beginning that effective enterprise computing and groupware don’t depend on transparency, replication, and semi-structured databases. They depend on how individuals are rewarded and punished for sharing and withholding information. They are about behavior, culture, and politics. Only with the appropriate incentives will employees use technology the way the organization wants it used. Information executives should not and must not try to change the culture on their own. They need to bring the human resources department and general managers into the implementation process as co-designers of the incentives and practices intended to encourage the desired uses. An appropriate incentive might be to insist that individuals whose work is cited, referenced, or downloaded the most receive cash awards or public recognition. Culture and technology must be managed ecologically. When the right cultural inducements are in place, organizations can decide on their collaborative information technology—and they should choose the technology that best fits their organization. Davenport’s book effectively explains the challenges information technologists face in helping to develop more productive corporate cultures. However, he could have devoted more space to the perverse impact information technologies can have on corporate relationships. For example, in many E-mail systems and intranets, users have the option of screening out the messages of certain individuals whose contributions they find useless or annoying. Are those filters a useful, timesaving network intervention or a potential source of political fireworks? Should management information system groups offer filters to their corporate clients? Should the filters be managed on the client or the server level? Should information services—or anyone in the organization—track who is filtering whom, or is such knowledge forbidden fruit? Maybe only senior managers should have the right to filter their subordinates. As the network becomes a major medium for corporate coordination and discourse, such questions become unavoidable. Such questions in turn raise other questions about who should have access to what information and at what times. Consider the use of software programs to help coordinate meetings and calendars on the corporate intranet. Who gets access to the calendars employees are now required to post? Can anyone view a calendar and request a meeting? And will everyone also be able to schedule meetings themselves? Can a manager block certain individuals in the organization from scheduling meetings? And can the boss override someone’s appointment and schedule a different meeting? Scheduling software threatens to make employees vulnerable to losing control of their own appointments. So when a manager logs on to a calendar, should he be surprised to find personal schedules filled with fake appointments intended to deter unwanted assignments? Digital deception may become the best way to protect your calendar. What happens when your information technology infrastructure creates powerful incentives for dishonesty? The ecology of information becomes the ecology of relationships. I suspect Strassmann would respect Davenport’s perspective but also be leery of any claims that information ecologies would be more productive than information systems. Perhaps a better approach than Davenport’s is to see that all these new technologies, all these choices, enable organizations to design their own nervous systems. How should intelligence and information be distributed? Where should the network be particularly sensitive or insensitive? Does the organization need one “backbone” or dozens of them? Asking difficult questions isn’t the challenge; making the effort to answer them honestly is. Technologists can no longer build and wire a brain while glibly insisting that the mind isn’t their responsibility. The rise of enterprise computing is finally forcing them—and top-level management—to realize that their networks depend on the culture and politics of their companies. They should design and manage accordingly.

The Real Problem with Computers

Partner Center