Governance for better data quality

So data analytics can make pumps speak. Machinery give information and this single version of the truth leads to predictive maintenance in manufacturing environments. In most organisations though, both big and small, quality is still a concern. “The reason for this is a lack of governance”, says Sophie Angenot, President of the Data Quality Association.

Governance is about formalizing existing processes, not only to improve data quality but also to comply with regulations. That is important to us all. We all have personally identifiable information in our databases, we all need to be in order with GDPR.
  • A first step in this formalization process is defining four roles: Identify your data producers: if you want to gather information, you must know its sources – be it machines (IoT)or people
  • Do the producers know the data consumers and their needs? They are not necessarily the same person and that’s a part of the problem of lacking data quality
  • Appoint a data owner to take strategic and budget decisions on governance
  • A data steward works on the data and sees opportunities to improve. It’s the most important person in your framework

Business and IT need to collaborate on this

“You need ownership, but the new model will inevitably be collaborative”, Sophie Angenot said at last month’s CIO Global Perspectives. “IT needs input from business. And that collaboration will only work in a well governed environment.”
Sophie concluded with a tip: “In order to help clarify ownership and the different roles, we often use Bain & Company’s RAPID Matrix.”

More information about the RAPID matrix is here.