Data Quality Measurement Matters

Mar 16, 2011 by Jim Harris in Data Quality

Recently on the Data Roundtable, there have been several great blog posts about measuring data quality written by David Loshin, Dylan Jones and Tom Redman. In this blog post, I decided to weigh in on this discussion about data quality measurement matters.

We have all heard the cliché: “If you can’t measure it, you can’t manage it.”

Well, just because you can measure it, doesn’t mean that you can manage it. And just because you can manage it, doesn’t mean anyone cares about it. For example, let’s use my height and weight.

I can measure my height (6 feet, 3 inches), but I can’t manage it since I can’t make myself any shorter or taller. I can measure my weight (200 pounds), and I can manage it since I can put myself on a diet and exercise program. I can create a baseline measurement before starting my weight management program, track my progress over time and create an executive dashboard for myself.

However, the key question is: Who, other than maybe myself, actually cares about how much I weigh?

I have witnessed a similar flaw with most data quality metrics. Some quality characteristic of data is identified, which can be both measured and managed. For example, the validity of a postal address. Postal validation software helps establish a baseline measurement as well as improves the validity of existing and newly created postal addresses, all of which can be easily tracked over time.

An executive dashboard is created and the data quality team proudly presents it to CxOs, who typically respond by asking the obvious question: “Why do I care about the validity of a postal address?”

Data quality measurement matters. However, the way you measure data quality impacts the way you manage it — and what must be measured and managed is the business impact of data quality.

When you measure the quality of data independent of how your organization is using it, then you aren’t providing insight into how data quality impacts business performance. When data quality measurements don’t provide this business insight, data quality measurements don’t matter.

Tags: Data Quality, measurement