Master data management —putting together the processes, governance and standards that turn your critical data into a single, reliable source of good and consistent, operational data —is not without its challenges.
As we said in our first blog, some of the typical stumbling blocks to a successful MDM implementation include taking a data-centric approach, for example failing to include the business relationship or process complexity, as well as not defining requirements clearly and not understanding the roles and responsibilities related to master data management.
But with so many moving parts and integrated functionality, how do you control the process of MDM and how do you consistently control and coordinate so many varying data quality tools within the MDM solution? What’s more, how do you successfully align this functionality and the data it manipulates to real-world business requirements?
We believe that the first driver in acquiring an MDM system should be tied to process improvement and we have identified three process-driven MDM pillars to achieve success:
- Ensure that your MDM is a business–driven discipline, supporting process optimization or transformation
- Determine that your MDM program scope is directly driven by process optimization needs, and MDM investments are tied to and measured by process improvement benefits and return on investment
- Make certain that your MDM follows a cross-disciplinary approach involving stakeholders from different functions or business areas which are impacted by the optimized process
A data governance “governing body or council” in your firm should be comprised of both technology and business specialists. But since data governance initiatives tend to start at the project level, as opposed to organizationally top-down or company-wide, a defined set of procedures needs to be repeatable and scalable.
It is the responsibility of the business to pinpoint how good and consistent data quality will improve process performance, help enable major initiatives by removing bad data bottlenecks and increase the bottom line. In turn, IT is tasked to find and evaluate the best technology platform for the job.
The MDM technology process comprises three major integrated phases:
- Data modeling and acquisition: Includes data modeling tools and the ability to import, format and pull data from all relevant sourcing systems
- Data governance: Includes excepting failed imports, monitoring external system connections, imposing authorization and stewardship controls, matching/cleansing, data-quality dashboard-ing and a workflow/approval process
- Deployment: Includes (upon approval) distribution of cleansed data back to systems of origination , beneficiary subscribing systems and research-accessible or business operational viewers (iPad®, intranet, portals or embedding for third-party applications)
Best practice would dictate there must be clearly planned and delineated process steps that drive MDM to create its most desired result: an enterprise-accepted single version of any multi-domain master data, code sets or hierarchies. Data quality should not be viewed as a random application add-on but a core component to business processes and their operational efficiency.
We have given you the fundamental steps to take to implement a process-driven MDM. In our next blog we will discuss the choices your organization has for implementing MDM—should you use pre-existing corporate data models for an MDM implementation, or choose domain-specific models/templates, provided by the vendor? Stay tuned.
For more information please download our free white paper: A Primer for Process-Driven Data Governance.