Degrees of data standardization - Part 3

As part of this series, in Part 1 and Part 2, I shared with you foundational activities necessary to prepare for your master data management program.

At this point we have a better understanding of our core data elements and their current state of quality - primarily completeness and accuracy. If you are working with a consulting partner, it is a good idea to ask them to share with you most common core elements they have seen at other companies within and across industries. Most of the consulting companies have reference data models supporting each industry. This gives your data organization an opportunity to synthesize and harmonize the data elements from across the organization and include any missing data elements that is required to complete master data. Slowly you will see the emergence of data groups and elements towards your new data model.

This is also a good time to begin thinking about ways to source the new data elements. I will address the sourcing of data from both internal and external data providers in a future blog. Note, how we are switching back and forth between managing structure of data and its content.

With profiling and interpretation of current state of your data, you have by now identified data elements that are candidates for standardization. Though it is tempting to standardize as many data elements as possible for consistency across the organization, I suggest you exercise caution. This also applies to managing hierarchies across business units, divisional and regional boundaries.

Data standardization challenge is not about technologies or processes. It is about people!

Consumers and managers of data have a strong proximity to data they currently use. It is a relationship that has evolved over years, if not decades. They have grown with this data. They have turned a blind eye to some of its problems. They have found ways to work around it by interpreting it differently or by making "minor" changes in presenting this data to others. Ask them to adopt a different set of data and all of a sudden you have a battle on your hands.

This is a perfect opportunity to bring the passionate owners of data together. I have been part of several global consensus workshops to drive common data hierarchies (sales hierarchies, product hierarchies to name a few). These workshops become forums for vociferous turf protection, they are extremely political, even "Me Vs You" exchanges and have all the ingredients of a battle that is not for the faint of heart.

You must use this opportunity to make data consensus a cornerstone of your master data future.

When sparks fly during data standardization exercises you cannot but help admire the positive side of the issue. Any opinions you had until now that data is not respected or there is nobody who cares about data and all other ramblings about lack of organizational focus on data ... begins to evaporate. You are convinced that data is not just a byproduct of your applications anymore. All your organization needed until now was a forum to bring people together to decide on what is best for the company and information evangelist(s) to guide them through this process.

Data standardization just gave you that forum.