Data Governance: making progress as complexity increases

Often organisations wrestling with data governance are concerned about awareness (ownership, culture, promotion of data governance) and implementation issues. With the onset of the new world in data (big data, big analytics, data scientists), data governance questions take on a whole new level of complexity.

  • Joanne Leseman
  • 27 juni 2013

Earlier this week I was listening to a presentation from the Chief Strategy Officer of Reed Elsevier. I was amazed with their strong focus on data, information, customer insights and how they were using these to transform their business. It got me thinking about the problems they face in terms of Data Governance due to the size and diversity of their data sources. As with many other organisations, rather than explicitly choosing to “do Big Data”, Big Data has found them. 

So I decided to examine the link between Big Data and Data Governance using the 4 focus areas of our Ordina Data Governance framework (Quality, Data Governance Strategy, Culture and Ownership, and Metadata) asking the question: how relevant is our Data Governance framework when it comes to Big Data?

Big Data Governance – what are we talking about?

A quick Google of Big Data Governance leads to some relevant points but also, surprisingly, to some very big gaps.

  • Big Data Quality

In terms of data quality, there is an important and clear link between Big Data and Data Governance. IBM call it the 4th V, Veracity, defined as “conformity to truth or fact; accuracy.” (http://www-01.ibm.com/software/data/bigdata/). When organisations use data there needs to be a level of correctness present but just how accurate the data needs to be depends on its purpose. How to achieve this balance however is a subject of much debate. 

  • Strategy, Culture and Ownership

For Data Governance Ownership and Culture, the data sources (big, small or somewhere in between) seem less relevant in determining if an organisation has a strong data culture. A data centric organisation that has made the move to big data will transfer their knowledge and  data centric focus to the new big data projects.

Interestingly enough, finding owners for big data seems less of a problem than finding employees willing to be responsible for data generated from existing applications. Perhaps this due to Big Data’s exciting image which everyone want to be part of it. However, frequently this leads to situations where the big data has no logical owner earlier in its lifecycle, leading to a whole other set of challenges.

  • Metadata

Thirdly we have Metadata. Bringing data sources together is difficult enough when the processes involved happen at a normal speed. How do we manage this effectively when we have many more data sources, much more data per source, and all of it happens very fast?

If we look at masterdata, can we use our existing masterdata to evaluate new big data sources to check their relevance? Or on the flipside, do we use big data as a new source of masterdata, to extend the dimensions we can use to effectively interpret our big data? That there is a link is clear.

Data lineage is just as important for big data, not just from quality and auditing standpoints but also in addressing potential issues for example credibility, relevance and privacy.

  • Data Governance Strategy

Oddly enough, when it comes to an organization’s Big Data Governance Strategy, there is almost nothing of note to find. Arguably the most important aspect of Big Data Governance and certainly the foundation of any efficient and focused Data Governance program, a Data Governance Strategy guiding necessary actions and mapped to the organization’s strategy is mentioned nowhere.

So where do we go from here?

The conclusion from all of this is that while the 4 focus areas of the Ordina Data Governance approach are clearly just as relevant for Big Data as for standard existing data streams within organisations, surprisingly little has been done up to now to develop the tools and models necessary to help organisations make the progress they need.

Watch this space as we keep you posted with our progress addressing these issues in Ordina!

Over de auteur:

Joanne Leseman

Joanne is business consultant bij Ordina VisionWorks. Ze houdt zich bezig met all-round Business Intelligence, Data Governance en Business Intelligence Competence Centres. Joanne is veelal werkzaam in de Finance sector en is nu actief bij Elsevier.