My father was a refrigeration engineer, a line man who didn't think much of MBA types in staff positions. He designed and built plants to freeze vegetables. He would rather be out in the field building systems and shaking them down than reading about the latest management fad. In fact, the only management guru I ever heard him refer to favorably was Peter Drucker, so it's probably no surprise that I'm also a Drucker fan. I think Drucker does a marvelous job of identifying important trends—not just fads—that will affect the way organizations do business. In Drucker's article "The Next Information Revolution" (Forbes, August 24, 1998), he observed that organizations are moving beyond the mere "collection, storage, transmission, analysis, and presentation of data" to understanding the meaning and purpose of internal data and of integrating it with data from outside the organization. He also said that most organizations are too insular and need to do better at locating that outside information and integrating it into the decision-making process. At the February 1998 Fortune IT Strategy Forum in Pebble Beach, California, Drucker told executives, "The single biggest challenge you face is to organize outside data, because change occurs from the outside." He also noted, "Management is swamped with inside data, but doesn't have any more real information than it did 40 years ago, and the quality of decisions has not improved."
Industry Week interviewed Drucker last September about the state of manufacturing. Drucker's observations merit consideration. He invites us to redefine manufacturing as "the systematic process of production" and to think of service industries in terms of how they manufacture (produce) their products. He suggests that we consider the processes of handling thousands (or millions) of credit-card transactions, insurance claims, and even customer-service inquiries, as production systems, not just services. Drucker's observations rang a bell with me when I thought about how we database professionals use the term production systems to refer to the OLTP or other mission-critical systems we oversee. We base these systems on databases that contain our organizations' crown jewels. Our jobs are to maintain and protect that data.
I think most of us realize that "data is job one," but we also think about our data as part of an information supply chain. Why not extend the analogy and consider the similarities between information supply chains and the supply chains that make up a traditional manufacturing process? Let's see whether we can learn anything from manufacturing. Here are some examples of innovations that contribute to manufacturing companies' success: automation (robotics and other technologies mean fewer workers are needed); quality (techniques such as the integrated use of statistical process control, quality audits, and aggressive solicitation of employee suggestions); simplification (reducing the number of suppliers); just-in-time delivery (minimizing the overhead associated with warehousing large quantities of inventory); flexibility (the ability to reconfigure assembly lines quickly); and better accounting (the ability to identify how much it costs to produce a widget).
Each of these innovations has analogies in database administration. Consider automation: Microsoft SQL Server 7.0 is a great example of a tool that helps automate data handling. Experiment with the new SQL Server Agent technology to make sure you're automating as many tasks as possible. Think about how your developers build applications based on SQL Server. You can automate some of the process, and code and generate diagrams, by using wizards or new tools such as Visual Studio's Visual Modeler, Visual Component Manager, and Microsoft Repository.
As for quality, when was the last time you asked users whether they trusted your data and whether they ever received conflicting answers to the same question from different systems? You can't go wrong if you ask users for their pet peeves or ideas for improving data delivery. Think about using simulation, forecasting, and other management tools that can help you track information delivery quality of service.
Regarding simplification, consider consolidating data sources and using data archives. For legacy data that you can't pull the plug on, have you at least wrapped parts of the key applications? The notion of just-in-time data delivery coincides not only with basic decisions about whether to replace tree-consuming paper reports with electronic versions, but also with decisions about push vs. pull distribution. And think about flexibility: Most of us have invested time and energy educating users to perform ad hoc queries, but now we need to focus on system design. We need to be flexible so we can turn on a dime in response to merger and acquisition activity, to meet a request for a new data mart, or to support the newest campaign to manage customer relationships.
Better accounting is a hard-to-meet objective because quantifying the competitive value of timely information isn't easy. But if you don't try, you might be replaced with someone who will.
Explore the Similarities
I invite you to probe the similarities between modern factories and your organization's information production systems. Take a few minutes to visualize data flows in terms of assembly lines. Simply buying and installing SQL Server 7.0 won't magically give you a better automated information supply chain, or better quality control, or smarter and simpler data access. You'll have to dig in to the new management features built into SQL Server and think about your organization's basic infrastructure and information architecture. And then you'll have to figure out how to transform cool ideas into action.