Business Glossary and the Enterprise Brain Drain

In the early years of coal mining, miners would take a canary down into the mines with them because the bird would succumb to toxic gases before the miners and give them an early warning if something was wrong.

In information management, the warning signs of a toxic project aren’t always quite as straightforward. How can you tell it is about to go off the rails? I remember some projects in the early days of my career where the business analysts would use gallows humor to talk about how the project could never be delivered on time. Sometimes it is the people doing the work who know before the project planners.

Today, my information management canary is quite simple: Does the organization know its data? When it comes to reporting, it kind of helps if you know what you are building with.

But understanding data is not as simple as it sounds. There is a lot of data across business applications, and not much of it is clearly defined. Things get further muddled as data is moved around and copied and renamed and the people who know about these changes move on. The documentation that is created gets stale with time or sinks into the enterprise quicksand of project folders and shared drives. Enterprise brain drain—the inevitable loss of business knowledge through attrition and outsourcing—means that over time you know less about your data and rules behind it.

However, there is a way to treat information as an asset and manage it for lasting value and keep that canary alive. The IBM Data Governance Unified Process—a roadmap for implementing processes and tools—can help virtually every organization manage information better. It has been developed by the IBM Governance Council and is supported by a range of IBM software.

The core IBM product for improved data governance is IBM Information Server—a suite of tools that delivers the main building blocks of the unified process as shown in this diagram:

This basic roadmap helps you define your data, understand the content and manage those definitions over time—which helps you retain business knowledge in a structured format and reduces the impact of the brain drain. These foundational capabilities are required for any safe and successful data-driven project. These should be the mandatory tools of the trade, like a doctor with a stethoscope or a ghostbuster with a proton pack and particle thrower. You just shouldn’t tackle difficult jobs without them.

This governance roadmap is easy to implement by using the IBM Information Server, a family of products that makes it easier to manage and integration data and metadata across systems and databases.

• InfoSphere Business Glossary for defining enterprise terms and definitions and retaining business knowledge
• InfoSphere Discovery for understanding data via cross-system analysis and profiling and knowing what is really in the data
• InfoSphere Metadata Workbench for linking metadata across products and running impact analysis and data lineage reports.

These products provide a foundation governing information management projects and assets. But the most important product within IBM Information Server—the one that I consider indispensible—is the Business Glossary.

Countless governance canaries have died because the most important information is in the brains of subject matter experts and data analysts who have moved on to greener pastures or are too busy fighting fires on other projects. Common warning signs include “our experts are too busy to help” and “the team who built that is gone.”

If an organization does not have an enterprise glossary and does not want to start one I hear alarm bells. The concept is simple: a single version of the truth for definitions of information with a steward registered to help with queries about that information. Authors and stewards keep the glossary content up to date with change management tools and workflow steps to control the quality.

Glossary is one of the simplest and most effective of the governance tools. A glossary can define information elements, business rules, derivation rules, system interface rules, data sensitivity rules, and organization policies. It provides an entry point into stewardship and data ownership by assigning a steward and reviewer to all entries, and it puts information into context by linking terms to each other and to information assets such as Intranet pages, documents and database columns.

The InfoSphere Business Glossary stores content safely in a database and makes all functions available through browser-based applications. It is easy to implement and use and it is one of the key tools for combating the constant brain drain that devalues your information assets.

Using Business Glossary is the first step in avoiding common project disasters caused by common problems: misunderstanding data, not building what the business asked for, not knowing where to find information, not having access to subject matter experts, and not being able to find the right historic documentation.

Get started with Information Server bundles

I have helped companies of all sizes get started with Business Glossary, from major banks to universities and not-for-profit organizations. IBM has a range of bundles and discounts to make it affordable for small implementations and massively scalable for large enterprises:

1. Information Server WorkGroup Edition: The WorkGroup license makes all the products more affordable and is popular with smaller organizations. It offers a huge discount with some license restrictions on size—for example, no more than 500 glossary users.
2. Business Information Exchange: This is a bundle that gives you all of the data governance brain drain tools on a silver platter. Business Glossary, InfoSphere Discovery and Metadata Workbench under a simplified server and user license that can manage metadata from many products in a central hub.
3. Information Server a la carte: If you want an enterprise platform without any restrictions, you can choose individual products under standard pricing, picking from not just the products mentioned but also FastTrack for mapping, Information Analzyer for data quality monitoring, DataStage for data integration, and QualityStage for cleansing.