|Sign-Up for Free Exclusive Services:||Portals|||||eNewsletters|||||Web Seminars|||||dataWarehouse.com|||||DM Review Magazine|
|Covering Business Intelligence, Integration & Analytics||Advanced Search|
DM Review welcomes Rich Cohen as an online columnist. Each month he will address different focal points of BI strategies. He will draw upon more than 27 years of experience in the design, development, implementation and support of IT in his columns. Watch for them the second Friday of each month.
If your company is like many others today, it's been cobbled together from a series of consolidations, reorganizations, buyouts, takeovers and mergers. Even if you haven't been hit by one of these corporate events, your information technology (IT) architecture has probably leveraged an assortment of homegrown legacy systems and packaged products that are ostensibly integrated - but often aren't. While mergers and growth may create synergies that didn't exist before, they can also create bottlenecks when trying to integrate information from various IT systems. The result? Data chaos.
Data chaos is a disorganized IT environment characterized by inconsistent, unreliable, redundant, and de-centralized data. Data chaos can paralyze your IT staff's ability to get management and knowledge workers the information they so desperately crave. In a chaotic data environment, data is housed in different systems, which often results in manual manipulation and adjustments to make data conform to managerial reporting and accounting standards.
To exacerbate the problem, corporate data packages often have different formats and content. What's more, the business units frequently have different account or line-item definitions as well as different time frames or definitions of time. Again, intensive manual effort is required to adjust, reconcile and centralize corporate data to reduce redundancy in data and reporting streams.
The result is that inconsistent data is often used as the basis for analysis and business decisions. Overall, the data manipulation and analysis process is cumbersome, and accurate forecasting and planning becomes nearly impossible. If you're in this situation, your options are 1) to put your hands over your ears and moan, or 2) to look at the chaos as an opportunity to fix the problems. Wouldn't you choose the latter?
First, let's take a look at the components of the ideal data management (DM) environment. Then, we'll map out a plan to get there. In the ideal DM environment, data is managed at the enterprise level. You think you're there already? Unfortunately, the reality is that while some companies truly do have an enterprise DM strategy in place, most of them rarely govern their data at the enterprise level.
There is a difference. What about the systems that feed the data warehouse or data marts? How is that data managed? To truly manage data at the enterprise level, you have to move beyond the data warehouse level to include ERPs, legacy systems, and even systems that don't directly feed the data warehouse or data marts. In other words, theDM environment must be enterprise-wide, and it must be integrated.
There are five integrated components that underpin a solid, enterprise-wide DM environment:
Figure 1 represents the components of an enterprise-wide, integrated DM environment.
Let's look at the components in depth. The first component, data standards, should be the starting point. Many companies start with the third component - the technology architecture - but they shouldn't. If you don't know what information you have and what format it is in, how can you select tools for managing it? To develop your organizational data standards, the first step is to build a data model containing common enterprise-wide entities, attributes and their interrelationships. Next, define records of authority for each data object and data element. This confirms that data is consistently defined across the enterprise and that each process and service using the data is "speaking the same language."
The second component is data processes. It is here that you define and deploy processes to confirm that enterprise data is properly entered and maintained, and that the standards defined in the first component are implemented. This second component also includes data synchronization processes to "publish" the enterprise-level data and allow for local data element maintenance. Finally, it also includes processes that are responsible for guiding the path of data throughout the organization, such as:
The third component is the technology architecture. This consists of the data warehouse, data stores and reporting methods and analytical tools that are used to retrieve, consolidate, synchronize and deliver data across the enterprise. These tools turn data into usable information that is presented in comprehensible, consistent reports that drive business analysis and strategic planning. They should be best of breed and standardized throughout the organization.
The fourth component is the data organization. This component deals with the human factor in the DM environment. This is where you design and implement a data management function that will have the responsibility and authority to govern enterprise data. It's also where you appoint and identify the organizational data stewards, key stakeholders and owners of data, along with their roles and responsibilities.
The fifth component is the governance architecture. The data organization formed in the fourth component will direct the development of the governance architecture. The architecture should include enterprise-wide policies and procedures, organizational structure plans, skills matrices and training requirements for the DM environment. The governance architecture should be the last thing you do when setting up your DM environment. Until you know what information you need and how you need to present that information for analysis, you cannot develop policies and procedures to govern it.
Once you have the components of an enterprise-wide, integrated DM environment in place, you can begin to reap the benefits:
The key here is to keep thinking at the enterprise level. Data is managed day to day in the trenches, but data management strategies are executed at the enterprise level. If you build each of your components for the DM environment with an enterprise view in mind, you can move beyond managing data in relation to the data warehouse. You can employ an enterprise-wide data management strategy that will enable your organization to leverage its data and not just survive, but thrive.
Rich Cohen is a principal in Deloitte Consulting LLP's Information Dynamics practice where he is responsible for the strategy, development and implementation of data governance, data warehousing, decision support and data mining engagements to support the emergence of world-class business intelligence applications. Cohen has more than 27 years of experience in the design, development, implementation and support of information technology in a variety of industries. Over the last 18 years, he has had extensive experience in the creation of technology strategies, implementations and deployment of CRM and business intelligence solutions to drive improved business performance.
|E-Mail This Column|