Data Quality Requires Data Governance

I’ve seen many firms throwing good money after bad at their data quality problems, cleaning the data over and over, but spending little effort getting to the root cause –poor governance over the data lifecycle.

Data quality starts in the top right quadrant in the graphic below. It starts with defining what quality requirements are, how the data is intended to be used, the criteria by which it will be measured and the standards and expectations for use. Then it requires the processes and resources to ensure that those things are understood and consistently followed – lower right quadrant.  

On the left of the diagram is data operations and the things we do to ensure the data is fit for purpose. Beating up the folks working on the left, without ensuring the governance program is functioning as it should (i.e. the things on the right), is frankly, not very smart.

If you have systemic data quality problems, look to the right (symbolically speaking) and ask, is my data governance program functioning as it should? If you have poor quality data you certainly have to fix that, but don’t just fix the data, fix the problem. More times than not, the circumstances that allowed the quality to falter is because of governance problems.  

Because of that last statement, it is critical that the governance program is complete, functional, and sustainable. We’ll talk about that next.