Top 14 Master Data Management Misconceptions
- It’s for the enterprise and enterprise requirements are impossible to get
While ideally it is for the “enterprise” (which may not be your entire organization), MDM still has to deliver to applications on its way to supporting multiple applications and, at some point, being able to truly be enterprise.
2. It’s for a single application so it doesn’t have to be a separate organizational discipline
MDM efforts should be leveraged beyond a single application or subject area. The work in building the MDM foundation has high leverage to enterprise application delivery over the course of the next decade at your shop. Don’t architect MDM so tightly with the first delivery that it is difficult to go beyond
3. It’s all about _________________
Fill in the blank with data quality, hierarchy management, merge/match processing, workflow/governance, real-time data integration, an enterprise data model or something else. In reality, it’s all of the above. While one or more of these value propositions may be most interesting to begin with and may kick off the project, know the full spectrum of possibilities for MDM and be ready to utilize each when needed
4. I can bring master data together in the data warehouse
Yes, you can, but a batch-loaded data warehouse is too late in the data lifecycle to be effective in real-time processing. Even a real-time data warehouse usually lacks the many functions of MDM
5. Most subject areas don’t need workflow/governance
Many don’t need elaborate workflow/governance because they are originated elsewhere to standard, and MDM will incorporate that data as master data. Often this is the case with Customer, as with a POS system. However, smaller workflows for authorization and enrichment add value even to this data
6. There’s no Return on Investment
Technically, this is correct. Unless MDM is considered part of a business application that delivers returns, MDM is all investment. The returns come in the improved efficacy of the projects that use the data, many of which can do their function – ultimately sales or expense reduction at some level – much better as a result of MDM data instead of whatever they’d do on their own. Enterprise MDM also delivers lower organizational Total Cost of Ownership by doing something once that will used over and over
7. Those projects all seem to fail
Failures capture attention and MDM seems to have caught the bad luck of having its failures broadcast. All projects fail if you do them wrong. MDM does require business input for success. It is not a strictly IT project and treating it this way will lead to failure. Well-done MDM is providing companies in all industries of all sizes very high value
8. I start my journey by choosing a MDM vendor
Start with quality vendor-neutral education and consulting. There will be time for vendor involvement later
9. I can avoid data modeling by doing MDM
The model is the most leveragable component of MDM. Everything you put into the data model will pay off many times over. You should customize packaged models and expect to put work cycles into this effort.
10. The best approach to data quality for MDM data is to start by figuring out a data profiling tool and running it blindly against my data
Do this if you enjoy unproductive work. Data quality is custom and efforts should start with understanding a profile of what rules the data should be conforming to.
11. Data Quality is nebulous and intangible
Most effectively, data quality is scored and very tangible.
12. All MDM tools are the same
All tools are not the same and some will not even do all MDM jobs, no matter how hard you try. Some provide a lot of intellectual property (models, reports, workflows) around subject areas and others do not. You should decide what’s important to you.
13. Organizational acceptance will take care of itself
Organizational acceptance is the hardest part. Include organizational change management in your MDM implementation for success.
14. Third-party data is inappropriate for MDM
Third-party data is largely about extending the profile of important subject areas, which are mastered in MDM. Taking third-party data into organizations has actually kicked off many MDM programs.
This post was written as part of the IBM for Midsize Business program, which provides midsize businesses with the tools, expertise and solutions they need to become engines of a smarter planet. I’ve been compensated to contribute to this program, but the opinions expressed in this post are my own and don’t necessarily represent IBM’s positions, strategies or opinions.