Behind the Curtain: Integrating Data and Analytics for Better Governance
In my last blog about Gartner’s Data and Analytics Summit, Putting the Data Citizen Front and Center, I recapped what I see as a sea-change in how data governance is being practiced. More and more, organizations are recognizing that value lies with giving business users the data they need to solve business problems rather than asking data analysts to search through reams of data without a clear understanding of business needs.
In this blog, I’d like to dig a little deeper to look at one way we might accomplish that. Let’s begin with the premise that business users need information to do their jobs–not just data. That is, they need access to the full spectrum of artifacts that support the production of knowledge, with data at one end and reports and analytics at the other. In his talk, Data and Analytics Governance: Coming Together, Gartner analyst Thomas Oestreich made the argument that artifacts as varied as data flows, algorithms, MapReduce jobs, and decision models (to name just a few) are just as important—perhaps more important—to data citizens as are raw data sets. But here’s the catch: just like data, which a business user will ignore if she doesn’t trust its quality, these data artifacts are useful only insofar as the data informing them is easy to see, understand, and trust. In other words, analytics need to be integrated into a data governance framework. Prioritize which analytics are valuable to the business, create agreement on how they are to be shared, make those artifacts easy to find, and (the secret sauce I would argue), make the linkages between analytics and data transparent.
The convergence of data and analytics in a governance framework is exactly what the Collibra Data Governance Platform is designed to do. Any object, from reports to MapReduce jobs, can be easily added to the governance environment. Once again, the business decides which analytic datasets provide the most value to the company. Governance, then, is a matter of applying the right processes to enable the same level of trust in more complex outputs like data portals or KPIs as any other data set. This is how organizations build trust in the data your business relies on and the analytics used to make decisions and solve business problems.