Master data management can be defined as the combination of people, process and technology necessary to manage non-transactional data for the organisation - most commonly (but not limited to) customer, supplier and product data. Unfortunately, the definition sold to business by vendors is frequently about “centralised facilities designed to hold master copies of shared entities, such as customer and product.” This repository-centric view of MDM ignores a critical area for its successful implementation - the data, and in many cases is a direct contributor towards MDM implementations failing to meet business goals.
According to the Butler Group review: “An Introduction to Master Data Management”, MDM is evolving as a solution to data quality issues within an organisation. Of particular importance is the fact that data are not static entities - even in well-managed organisations, data will deteriorate or become stale over time. The review continues to suggest that among the critical tools for supporting MDM are data quality solutions for data discovery and profiling, and matching, cleansing and enrichment.
Data quality provides assurances that the integrated data is consistent, complete, and fit to publish to the business community. The Butler Group reviews MDM solutions from a number of vendors, including data quality vendor, Harte-Hanks Trillium Software.
According to the review an investment in data quality solutions:
* Accelerates the MDM project implementation process (by re-using cleansing and matching rules developed over 20 years);
* Optimises the investment in MDM by enhancing the validity of data;
* Reduces the need for manual intervention through more accurate matching; and
* Provides a lower total cost of ownership due to continuous updates and improvements to the global country rules.
The review also discussed the importance of a pragmatic data governance programme and metadata library to manage the process, define the domains, and measure the compliance of data to the business rules. Data governance, with the support of a data profiling tool, allows you to establish the metrics to identify where your MDM program may be struggling, and put in place the necessary data management processes to address these potential points of failure.
The use of an automated data profiling tool during the data discovery phase will help to identify project risks due to incomplete or inaccurate data documentation. Project teams that identify and manage these risks early save considerable time and costs compared to those that wait until the test phase of the implementation to discover that the data does not fit the new data model!
Share