Chalk Talk: Business Driven Data Governance for Analytics
Welcome back to our series of Analytics Chalk Talk videos. I hope you enjoyed our last video: “Driving Analytics Data Management Competence with Metadata.” In this Chalk Talk, I’m discussing how organizations can develop better, more business-driven approaches to data governance. The result will be great data that will deliver valuable analytics insights.
Please watch the video or read the transcript. You might be able to tell I’m pretty passionate about this particular area, so share your views in the comments below or at Twitter–@leanlyle.
With the increase in sophistication of our analytics and analyses comes the realization that the quality of the data and agreements on its governance are no longer nice-to-have, but essential.
We know data governance programs have a tendency to collapse under the weight of undefined roles, steering committees, and data stewards. The better, business-driven approach is to start small and build on success. Gradually incorporate those successes into the business and you’ll create a self-sustaining culture.
But what does that look like in practice?
Start by picking a specific problem that governance can help with—something where bad data is breaking a process and costing an identifiable amount in dollars. That way you can engage the business. For example, the cost of each product breakage multiplied by the number of bad records gives you a baseline for the cost of that poor data quality. Or choose a compliance problem where you’ll be fined for having inaccurate data.
As a working example, Cleveland Clinic wanted to be able to analyze their data in order to do predictive demand forecasting for over 100 operating rooms, plus everything downstream, with two months’ lead time. You can imagine the volume and variety of data from different sources they had to work with and have agreement on. But they did exactly the thing I’m suggesting: they identified the different data pieces they needed to improve, engaged the business and marketed each success.
Successful data governance programs start small and build to become self-sustaining over time. The steering committee and data stewards come later. This only becomes more important when we’re dealing with hybrid environments, cloud and big data because the world and our analysis requirements are becoming more complex and sophisticated all the time. This is why data governance is no longer something that’s nice to have but something we have to have.
Other videos in the series: