There is a huge difference between taking a snap shot measure of something and measuring something over time. This is nothing new, but the famous management guru Peter Drucker can take credit for introducing and formulating measurements over time to truly improve business processes.
It’s worth sharing Peter Drucker’s contribution here, because he played a big part in helping Japanese businesses to mass produce quality products in the post-war era. His concepts for mass production made Japanese leaders regard him as legendary and sought his advice to modernise their corporations.
From his many contributions to management thinking, the measuring of processes over time as a way of increasing product quality became the basis of significant product improvements.
The idea is simple, just measure something you want to improve, measure it over a regular time period and use simple statistical techniques to ask poignant questions about the process and how to improve it.
Organisations found that they could measure a large number of important attributes, and by following his technique, they quickly made significant improvements.
So how does this relate to data?
I have mentioned on many occasions how an organisation’s data quality is reflected in the processes surrounding data but more importantly the attitude towards data quality.
The quality of data always reflects the quality of the processes.
Measuring data quality is still in its infancy, and if we only measure data quality every now and then, we are not likely to see much benefit, hence probably why data measurements are not yet mainstream.
The truly advancing organisation must measure their data quality over time. We’ve spoken about using data dimensions as one method (you can download our whitepaper below); defining your own metrics that are important for your business is a great first step.
Clean the data as best you can and then take the first snap shot using your metrics. This is the baseline measure of your data quality.
As you take repeated measures (could be daily, weekly or monthly) you will find trends, dips and peaks on your time graphs. Each of these events tells you something significant about the way data is being stored. Use this vital information to ask the right questions and you will soon find yourself as an elite organisation and creating better results based on reliable data.
How can I get started?
Feel free to download the Seven Dimensions of Data Quality, or read some of the previous posts. Call us if you want to dive into the details and start the process on sustainable data quality improvements.