Be Data Literate – Know What to Know

by Peter F. Drucker

            Executives have become computer literate. The younger ones, especially, know more about the way the computer works than they know about the mechanics of the automobile or the telephone. But not many executives are information-literate. They know how to get data. But most still have to learn how to use data.

            Few executives yet know how to ask: What information do I need to do my job? When do I need it? In what form? And from whom should I be getting it? Fewer still ask: what new tasks can I tackle now that I get all these data? Which tasks should I do differently? Practically no one asks: What information do I owe? To whom? When? In what form?

            A “database,” no matter how copious, is not information. It is information’s ore. For raw material to become information, it must be organized for a task, directed toward specific performance, applied to a decision. Raw material cannot do that itself. Nor can information specialists. They can cajole their customers, the data users. They can advise, demonstrate, teach. But they can no more manage data for users than a personnel department can take over the management of the people who work with an executive.

The First Challenge

            Information specialists are toolmakers. The data users, whether executive or professional, have to decide what information to use, what to use it for and how to use it. They have to make themselves information literate. This is the first challenge facing information users now that executives have become computer-literate.

            But the organization also has to become information-literate. It also needs to learn to ask: What information do we need in this company? When do we need it? In what form? And where do we get it? So far, such questions are being asked primarily by the military, and even there mainly for tactical, day-to-day decision. In business such question have been asked only by a few multinationals, foremost among them the Anglo-Dutch Unilever, a few oil companies such as Shell, and the large Japanese trading companies.

            The moment these questions are asked: it becomes clear that the information a business most depends on is available, if at all, only in primitive and disorganized from. For what a business needs the most for its decisions – especially its strategic ones – are data about what goes on outside of it. It is only outside the business where there are results, opportunities and threats.

            So far, the only data from the outside that have been integrated into most companies information systems and into their decision-making process are day-to-day market data: what existing customers buy, where they buy, how they buy. Few businesses have tried to get information about their noncustomers, let alone have integrated such information into their databases. Yet no matter how powerful a company is in its industry or market, noncustomers almost always outnumber customers.

            American department stores had a very large customer base, perhaps 30% of the middle-class market, and they had for more information about their own customers than any other industry. Yet their failure to pay attention to the 70% who were not customers largely explains why they are today in a severe crisis. Their noncustomers increasingly were the young affluent, double-earner families who were the growth market of the 1980s.

            The commercial banks, for all their copious statistics about their customers, similarly did not realize until very late that more and more of their potential customers had become noncustomers. Many had turned to commercial paper to finance themselves instead of borrowing from the banks.

            When it comes to nonmarket information – demographics: the behavior and plans of actual and potential competitors; technology; economics; the shifts signaling foreign-exchange fluctuations to come and capital movements – there are either no data at all or only the broadest of generalizations. Few attempts have been made to think through the bearing that such information has on the company’s decisions. How to obtain these date; how to test them; how to put them together with the existing information system to make them effective in a company’s decision process – this is the second major challenge facing information users today.

            It needs to be tackled soon. Companies today rely for their decisions either on inside data such as cost or on untested assumptions about the outside. In either case they are trying to fly on one wing.

            Finally, the most difficult of the new challenges: We will have to bring together the two information systems that businesses now run side by side – computer-based data processing and the accounting system. At lease we will have to make the two compatible.

            People usually consider accounting to be “financial.” But that is valid only for the part, going back 700 years, that deals with assets, liabilities and cash flows; it is only a small part of modern accounting. Most of accounting deals with operations rather than with finance, and for operational accounting money is simply a notation and the language in which to express nonmonetary events. Indeed, accounting is being shaken to its very roots by reform movements aimed at moving it away from being financial and toward becoming operational.

            There is the new “ transactional” accounting that attempts to relate operations to their expected results. There are attempts to change asset values from historical cost to estimate of expected future returns. Accounting has become the most intellectually challenging area in the field of management, and the most turbulent one. All these new accounting theories aim at turning accounting data into information for management decision-making. In other words, they share the goals of computer-based data processing.

            Today these two information systems operate in isolation from each other. They do not even compete, as a rule. In the business schools we keep the two apart with separate departments of accounting and of computer science, and separate degrees in each.

            The practitioners have different backgrounds, different values, different career ladders. They work in different departments and for different bosses. There is a “chief information officer” for computer-based data processing, usually with a background in computer technology. Accounting typically reports to a “chief financial officer,” often with a background in financing the company and in managing its money. Neither boss, in other words is information-focused as a rule.

            The two systems increasingly overlap. They also increasingly come up with what look like conflicting – at least incompatible – date about the same event; for the two look at the same event quite differently. Till now this has created little confusion. Companies tended to pay attention to what their accountants told them and to disregard the data of their information system, at least for top-management decisions. But this is changing as computer-literate executives are moving into decision-making positions.

Up for Grabs

            One development can be considered highly probable: Managing money – what we now call the “treasury function” – will be divorced from accounting (that is, from its information component) and will be set up, staffed and run separately. How we will otherwise manage the two information systems is up for grabs. But that we will bring them together within the next 10 years, or at least sort out which system does what, can be predicted.

            Computer people still are concerned with greater speed and bigger memories. But the challenges increasingly will not be technical, but to convert data into usable information that is actually being used.  (This article appeared in the Wall Street Journal, December 1, 1992)

Mr. Drucker was professor of social sciences and management at the Claremont Graduate School in California until he died in 2005.