When we discuss Process-Driven MDM, what processes are we talking about? Process_Gears

Well, for starters, certainly business processes. We understand that poor data quality is bad for business health and that healthy data improves the quality of business processes. Therefore, many organizations have come around to realizing that when implementing a master data management tool, it makes good, business sense to align process improvement initiatives with fixing that process’ master data.

At a high level, consider the mapping below of retail improvement projects to their respective data sets:

Retail Drivers for Process Improvement

(Hint: You’ll receive a flunking grade in Process-Driven MDM 101 if you decide to improve the quality of customer loyalty programs by addressing the mastering of the employee domain. Conversely, you should absolutely flunk Business 101 if you fail to address data quality as part of corporate process improvement or re-engineering).

MDM as Process

But we should treat the MDM tool (or more accurately, the MDM platform), as its own process.

Actually, more as a series of processes. When viewed as a series of multiple technologies, an MDM solution can be viewed holistically as a solution comprised of three major, integrated parts or phases:

I. Data modeling and acquisition: Includes data modeling and the ability to import, format and pull data from all relevant sourcing systems for cleansing and reconciliation.

II. Data Governance: Includes rejecting and queuing failed imports, monitoring external system connections, imposing authorization and stewardship controls, matching/cleansing, data quality dash-boarding and a workflow/approval process.

III. Deployment: Includes (upon approval) distribution of cleansed data back to systems of origination, as well as beneficiary subscribing systems and research-accessible or business operational viewers (iPads, intranet, portals or embedding for 3rd party applications).

Let’s drill down further in II, and view MDM’s workflow/approval function as its own technology process:

  • Workflow rules are configured to create approval routes based on specific events or event/attribute combinations.
  • Data-driven workflow rules route the approvals based on the data value(s) or type of change involved.
  • Data changes trigger internal alerts and notification through corporate email.
  • Additional rules in terms of mandatory attributes can be required (product data might require an upload jpg).
  • Best practices dictate that data be validated and approved by business owners

Now, let’s reconsider the same workflow functionality from the perspective of a business challenge:

Step 1: The Data Management Business Requirement: Fictional Stellar Applications, inc. decides to move its entire sales department from Philadelphia, Pennsylvania to corporate headquarters in Yonkers, New York.

Step 2: A data manager, filtering on all sales people in Philadelphia, performs a bulk-update in the MDM hub, changing all instances of city and state fields from Philadelphia to Yonkers, and Pennsylvania to New York, respectively.

Step 3: This city and state field change triggers a business rule for the workflow/approval process which notifies HR data stewards to review and validate the initial employee changes.

Step4: Since it is unrealistic to expect that all sales personnel in Philadelphia will choose to commute or live closer to Yonkers, HR must validate the employee’s work-status. if the employee has agreed to work in Yonkers, the bulk-update will be approved by the HR stewards. If any employee decides to resign from the organization, the change is rejected with explanation. But, that employee record will still be updated with a scheduled termination date, allowing that employee particular record to be retired or deactivated at the appropriate time.

Step5: All approved and governed employee data is subsequently scheduled by the MDM system administrator to be deployed to all relevant, downstream systems and databases.

Though extremely important from an HR perspective, it's a simple but effective example of sharing data management responsibilities between the business and IT.

By the way, if you'd like to view a short MDM video demonstrating the above workflow scenario as executed in webMethods OneData - - please click here.

 

Leave a comment

Search B2B

Most Popular Blog Posts

Subscribe

Connect with us!