Why MDM with EDW?
Common Business Objective - Enterprise Data Vision: Increasingly, enterprises are realizing the value of data and are using solutions such as MDM and EDW to help implement their enterprise data vision. Most large enterprises face similar challenges around data - the absence of an enterprise wide data strategy, predominance of application-based architectures, vertical or departmental-based data silos and absence of strong data governance. These data-related issues create redundancy and unreliable data repositories in the enterprise. Businesses are recognizing that these issues can lead to bottom-line inefficiencies as well as hurdles for top-line growth. This recognition is driving a data vision that aims to provide a "single version of truth" for the enterprise. An EDW enables this vision by providing data in one central place that can be accessed by multiple lines of businesses, processes and applications, providing consistent and accurate representation of data. MDM solutions are an integral part of the EDW in providing a robust set of technologies and processes to achieve the single version of truth.
Figure 1 : Enterprise Data Vision and Supporting Components
For example, consider a large bank that has multiple loan processing systems across the lines of businesses. Without an enterprise data warehouse, the same customer who opens a checking account, a mortgage account, a brokerage account and a commercial line of credit would be tagged with disparate identifiers. Further without an MDM solution, this results in an enterprise having a fragmented view of the customer. If a marketing campaign is based on customer behavior exhibited in the life of a mortgage loan and does not account for behaviors in other lines of businesses, it may result in lost opportunities. If customer profitability analysis is conducted based on behavior on consumer loans and not capturing behavior in the brokerage accounts, one would get an incomplete picture about the customer.
A data warehouse can be the first time when various business units that have worked as silos pause to think about the benefits from a common data infrastructure. It is also potentially the first time the pain points are uncovered during integration to achieve a robust enterprise data vision. An MDM solution can ease the pain of integration providing a common and uniform set of master data elements such as product, customer, organizational hierarchy, etc.
Figure 2
Common Supporting Components: In the previous section, we explored how MDM and EDW work toward a common business vision. Along with providing a single source of truth, the solutions can use the common set of enablers that drive successful implementations of both EDW and MDM solutions. A robust data strategy consisting of an overall data architecture and components such as data quality, metadata, data standards, data governance, data integration and data modeling are essential to building an EDW and as well as MDM solution with the EDW. As such, both solutions can utilize the common set of enablers to build a comprehensive solution. Figure 3 provides a list of features of these applicable to EDW and MDM solutions.
Figure 3
Data Governance: Data governance is a crucial and critical component that can strongly influence the success of complex projects such as EDW and MDM. The term governance is defined as the use of institutions, structures of authority and even collaboration to allocate resources and coordinate or control activity in society or the economy [Bell, 2002] . Thus data governance within an organization is the use of organizational structures, procedures and processes to control or manage data used in an enterprise. The organizational structure typically consists of a governing body of key business and IT executives along with a variety of other staffing resources for executing the processes. Note that this is a management function that needs strong executive support for it to succeed. In conjunction with a data governance program, a data stewardship function is usefully in assigning roles and responsibilities within the organization. Many enterprises are now in the process of implementing comprehensive data governance and stewardship functions as part of their data strategy.
The functioning of data governance in context with MDM and EDW could be best illustrated by the following examples:
Master data creation/migration: During the initial MDM solution deployment, standardization and harmonization across various legacy data systems is needed. Figure 4 shows a sample flow for resolving master or reference data issue. For example, consider the issue that customer names and other identifying information is not consistent from one of the application systems that feeds in to the data warehouse. The issue analysis may point to a customer facing application that is not designed to capture first name, last name and other identifying information (such as Social Security Number) in a consistent and unambiguous fashion. This may have to be fixed at the upstream application level to ensure quality data available as part of the MDM solution. The decision has to be taken by someone who has the authority to enforce these decisions and this has to be communicated across the various organizational units.
