One of the elements of a good data governance plan is establishing data quality metrics. Put another way, what are your measurements for how good your data really is?

One of the simplest but perhaps most powerful metrics is the ratio of data to errors (or what percentage of your data is correct). Simply put, you take the total number of a set of data and compare that to the number of errors on the list. For example, a committee list of 24 names and emails that has two errors on it would have a ratio of 24:2 (or 92% accuracy, if you prefer percentages).

The reason I like this simple formula is that it allows you to have an objective measure of data accuracy. Too often I hear from my clients "Our data is garbage" but they can't really quantify what "garbage" means or what data that is "not garbage" looks like.

There is a tendency to believe the data should be perfect. This is impossible, of course, as I've written many times over the years. But using a ratio of data to errors can help you quantify how good or bad your data is, and also help you set a measurable target for how good your data should be.