A panel of data management gurus shared their thoughts at a conference in London on November 9, 2005. This article is a summary of a discussion on the future of data management that included the author, Michael Scofield; Michael Brackett, a specialist in data resource design and remodeling; Alec Sharp, an author with expertise in process redesign; and Graeme Simsion, senior fellow, University of Melbourne and co-author of the book, Data Modeling Essentials. Many of these gentlemen will be presenters at the DAMA International conference to be held April 23-27 in Denver, Colorado.

The original topic posed to this panel was the future of data management (DM). However, I want to assert that the future of DM is somewhat irrelevant if you are not involved. Hence, I would like to reword the question to talk about your future in data management and in your enterprise.

To ensure your success in crafting the future of data management in your enterprise, you need to be entrepreneurial and proactive. You may need to change your behavior and expand or alter the focus of your job. At minimum, you need to:

  • Know your business,
  • Know your enterprise's data,
  • Give up idealism in data management,
  • Engage the organization and its people, and
  • Pick your battles.

Let us go through these points in more detail.

Know Your Business

IT and data management would not exist without the business paying the bill; hence, you must be curious about the business. You must know:

  • Who the executives are,
  • Where they came from,
  • The values and culture of the business and the industry(ies) in which it plays,
  • Where the business is going,
  • Who the principle customers are,
  • What constrains growth, and
  • The various environments (including government regulation) of the business.

Because some of the business's own behavior may be inferred from its data, you in data management should know your enterprise's data and what it can tell you (and your executives) about the business.
As part of understanding the culture of the enterprise, build a corporate glossary of terms, abbreviations and acronyms that are unique to the organization or its industry. A newcomer to the organization is in an advantageous position to notice unusual word usages. Even if you don't know the definition, write it down. Build up the glossary and seek out definitions. Both the development and delivery of such a glossary (and any other unique visual aid or map of the business and its processes) offer opportunities to interact with people all across the enterprise.

In addition to knowing your business, you need to know about the national and world economy at large, and how your industry and business fits into those.

Know Your Enterprise's Data

You cannot manage what you cannot see. You must have read-only access to all the significant data - a task which may require some time and political effort. You should strive to be the expert on the enterprise's data, not just the enterprise's metadata. You should know:

  • What data exists,
  • Where it is,
  • How to get at it,
  • Who thinks they "own" it,
  • Where it comes from,
  • What it means, and
  • What its quality is.

The scope of the enterprise data asset includes formal production databases, departmental and personal data of significance, unstructured data of significance, enterprise-owned data sitting on someone else's computer (such as a service bureau), and major imports and exports of data (i.e., data flows).
Along with all this, you should know the data and information needs (and wants) of your enterprise and its key decision-makers. To fulfill this, and to really manage the data asset, you should be looking for new external sources of data that would be useful.

Give Up Idealism in Data Management

Managing an enterprise's data is not an academic exercise. It is done to serve real needs, many of them immediate. You must be pragmatic.

Twenty years ago, DAMA folk talked about the "corporate data model." Forget it. Most large, mature enterprises have fragmented logical data architectures, and no single model will describe it all. Most enterprise data architectures are morphing toward ever greater complexity. Any model will soon be obsolete, so you must minimize the time invested in building one and focus your efforts on keeping your models up-to-date for the parts of the logical architecture of the business that are most important.

Understand when anything is "good enough"- be it a data model or the quality of a portion of the data asset. You will never achieve 100 percent perfection, and as you approach 100 percent, you drive up your costs. While one can define objective criteria of high quality data, how clean you actually should make the data depends upon user needs.

You must tolerate change and chaos in the organization. Indeed, seek it out. Where the enterprise is morphing most severely is probably both most interesting and most in need of your insights into architecture. Those are good places to use your mapping skills (both process and data).

Engage the Organization and its People

Many IT folk are introverts. They would rather focus upon technology than the organization and its people. "Just go away, leave me alone and let me write great code!"

Data management experts may be slightly less introverted, but they still need to know when they are instinctively avoiding contact with the people and their issues within the enterprise.

Get out of your cubicle or office. Be proactive and entrepreneurial. Establish relationships all over the enterprise. Prowl the halls. Develop gifts (such as a corporate glossary previously mentioned) which you can offer to get relationships going. Find out what people are doing and what their data needs are. This proaction extends far beyond IT.

Learn to communicate through many forms and media. Learn to match your expression (communication vehicles) with the audience and need. For example, a data model (or E-R diagram covering a whole wall) may be a repository of knowledge about the logical data architecture of the enterprise, but it is not at all the best expression of that knowledge. That knowledge may have to be broken down and repackaged for delivery in palatable chunks.

To communicate effectively, you must master all the tools available - text editors, slide shows, animation and beyond. You must understand the difference between:

  • Knowledge and expression,
  • Architecture and model,
  • The terrain and the map, and
  • Legacy, to-be and package architectures.

Do not think in PowerPoint, which is a mere vehicle of expression of ideas that are hopefully recorded in greater detail on some other "canvas" (such as a Microsoft Word document). But be ready to communicate in whatever mode of expression is appropriate to the opportunity. This includes "elevator speeches" - very short (under 90 seconds) verbal expressions of important concepts, such as:

  • What you do,
  • What is data management,
  • What is data architecture, and
  • Why data management is important.

You must deliver value to the organization constantly and immediately. Do not make promises to deliver some value three years from now. Immediate value may come in many forms. Don't be limited by your formal job description. You are an expert on the enterprise's data and what it says about reality. Leverage that to the benefit of everyone in the organization. If an office secretary is having trouble with Excel, help her out. Solve her problem immediately. Learn to be a good coach. Make her so happy she gives you chocolate.
Know your executives and what worries them. Anticipate their information needs by being very conversant in the issues they are facing.

Deliver value quickly. If an executive needs some information and you can write a simple SQL query against corporate data to answer the question, do it! Do it fast. Deliver it in a mature package (i.e., the information, perhaps a graph and table, couched in a document that explains the filters you used). Don't pass the request off to a data warehousing team who doesn't want to make eye contact with any business users. Deliver value now. If it becomes repetitive, pass it off to the reporting people later.

Pick Your Battles

If you are in a large, complex bureaucracy, you cannot do everything. You may have to be selective in where you deliver value. You may have to give up on some battles. For example, if a project is on the road to failure, let it fail. Don't get involved. Don't irritate the people if they don't want you. As someone once said, "Don't try to teach a pig to sing. It is a waste of your time, and it annoys the pig."

Look for easy wins. Look for quick wins. And when you do win, communicate your success. Make sure everyone in the enterprise knows the value of data and information, and that you are the person with the expertise about the data asset.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access