Home > Articles > Software Development & Management

The Information in Today's Organization

  • Print
  • + Share This
Although we generally can control information that we create, controlling and managing are two different processes. In this article, Adrienne Tannenbaum takes a look at how information is treated in today's organizations.
This article is excerpted from the book Metadata Solutions (Addison Wesley Longman, 2001), by Adrienne Tannenbaum.
From the author of

In this article, we focus specifically on the information over which today's organization has control. Usually we can control information that we create. Controlling and managing are two different processes, however. Ideally, we should control our information gracefully, and always for the benefit of those who need it most.

Information is often "managed" without regard to "today"—today being not a date, but a condition. It represents the state of information not only in terms of its actual availability, accuracy, and accessibility, but also in terms of its perceived characteristics. Because "today" is the result of "yesterday," many information management professionals are often quite focused on establishing a framework that will bypass today and get us to the "future." Let's look at how information is treated in today's organizations.

Information in Practice

The biggest disease affecting information practice today is tunnel vision. Many of us readily admit to tunnel vision in the past, but we hesitate to realize that we blindly follow the same philosophy and practices at the beginning of the third millennium. Despite reengineered business processes, streamlined application development, Web-based electronic commerce, and business-to-business partnerships, we don't seem to see the commonality of the information that spans these efforts. Worse, we never consider that locating and identifying information is very often more important than interpreting it. Obviously, if we can't find and identify information, we can't use it.

We are organized and often reimbursed based on the "tunnel" with which we are associated (for example, accounting, product development, customer). It is only natural, therefore, that supporting information is in line with that perspective. As Figure 1 illustrates quite simply, information tunnels are typically based on processing requirements. Even though the same data may be required in more than one tunnel, it is usually not easy to go from one tunnel to another. In any case, data is typically converted to tunnel-resident information, reflective of a limited set of requirements.

Figure 1 Information tunnels.

Organizations vary in their awareness of today's information situation. For the most part, upper management and executives rarely have to deal with the issues of information redundancy and integration. In fact, many large organizations have "information dissemination" functions created solely to respond to information requests from executives. People spend "as long as it takes," working nonstop if a deadline hangs over them to combine, interpret, sort, extract, and present the information known to address "enterprise" questions such as, "Who are our five largest customers?" In any organization with multiple billing systems, answering this question (if it is answerable!), for example, could be a major task involving the combination and comparison of sales facts that might contain conflicting or redundant customer identification. To prevent the constant repetition of the work required to massage this information, many data warehouses and data marts have been built to satisfy the need for information to support executive decisions.

The simple question about customer size that came from a consulting client more than 10 years ago could not be answered without a major overhaul of the client's data. The company, a chemical compound manufacturer, sold chemicals at various stages of production and mixed them with compounds that were often processed to create finished goods. As a result, many customers were also resellers, distributors, and manufacturing partners, depending, of course, on what was purchased and its chemical state at the time of purchase, as represented on the invoice. It was virtually impossible to identify the company's five largest customers without a down-to-earth data modeling session and a redesign of the company's various databases.

Many companies have survived or are currently living through a similar crisis. The dilemma of which many organizations are unaware is that they are often nowhere near as sure as they should be of the exact location, span, and penetration of the information they are seeking.

Figure 2 depicts a typical manufacturing scenario. Customers are identified inconsistently across applications, and some customers are also manufacturing partners. Given the information as it stands, how would we surmise who the customers are?

Figure 2 A conflicting and redundant customer/product information scenario.

  • + Share This
  • 🔖 Save To Your Account