Data Quality: The First Step on the Path to Master Data Management

Data quality is the critical first-step of MDM–without high-quality data, no MDM solution can be
expected to deliver value to the business. Without the data quality processes firmly in place, an
MDM solution will not only deliver poorer results than expected but could also end up costing an
organization dearly in terms of repairing the processes that lead to data problems.

The success or failure of any MDM implementation project hangs in the balance until both
business and IT stakeholders fully understand the issues and challenges for that project. On too
many projects, the business stakeholders are left to worry about business processes and change
management issues, while data quality, integration, consolidation, and migration are left to IT to
resolve and implement.

Exposing the data quality and data alignment issues and challenges early in a project, empowers
the business stakeholders to play a central role in the decision processes for the resolution of
these data challenges and enables the project team to thoroughly address the key risks that can
lead to project failure.

Just as data quality is the first step for MDM, data profiling (or more specifically a data quality
audit and scorecard) is the first step for data quality. During a data quality audit you should
measure the level and nature of data quality problems across multiple data sources -- to identify,
quantify, and categorize current and potential data quality issues and adherence to business
rules.

The audit should provide answers to questions such as “How bad is my data?” and “How much
impact is it having on business performance?” Knowing the size of the problem provides the
organization with a starting point from which to kick-off data quality enhancement and MDM. But
measurement doesn’t stop there; the audit process needs to be ongoing to quantify data quality
improvement over time and indicate areas where more work can be done.

Business-Relevant Data Quality Metrics

To ensure your organization has a proper understanding of data quality levels for key attributes in
its source systems, business and IT must collaborate to define a common set of metrics to
classify data quality and to describe defects and issues. This will enable better communications
between business stakeholders and the IT project team when describing issues or cleansing
processes and standards. The same metrics should form the basis for auditing and monitoring
data quality before, during, and after the MDM project has been implemented.

The specific attributes you choose to measure data quality will depend on what your organization
is trying to achieve and may depend on the stage of your project. For example, understanding
the completeness, consistency, and conformity of the data within and across each system is
critical at an early stage in the project. This will give visibility to the key issues and challenges to
be faced and the business users within the organization who must be engaged and consulted in
this process.