Eliminating the Barriers to Data Democracy

There's a lot of talk these days about how to streamline the data supply chain. And the discussions often boil down to how to control an organization's data and how difficult and time consuming it is for business users to access it. As I wrote recently for DataInformed, highly structured systems for managing data like master data management (MDM) and enterprise data warehouses (EDWs) put a kink in the data supply chain. They aspire to a single version of the truth but at a cost in time-to-insight few enterprises can afford to pay.

Complex Data Ecosystems

Large organizations also have increasingly complex data ecosystems that include plenty of application specific data silos, file shares full of semi-structured and unstructured information, NoSQL databases, Hadoop clusters, and streaming data. Hadoop alone has given rise to multiple data processing frameworks designed for specific analytical purposes.

Companies that have staked their future success on Big Data and analytics put enormous pressure on IT in its data gatekeeper role. They've got their nose to the Big Data window and they want what's on the other side. Now.

Of course, over the long run, everyone wants the same thing—success for the business. But the process of getting there often puts job-specific agendas in conflict. While LOB units worry about time to market and capitalizing on fleeting, data-driven opportunities, IT frets over runaway data growth, compliance, privacy, and data security.

Data Democracy, Not Anarchy

Truth be told, business users don't want to control an organization's data. They just want an easy way to see what's available and a simple method to get the data they need to solve a business problem. Control and access are related, but they're not the same. When we shop online, we have access to products, but we don't control the inventory. Nor do we want to. That would lead to anarchy.

The ideal solution would be a single access point where users can see all the data, without moving it. That way IT admins can make smart choices about where data resides and the governance protocols that control it.

As I discuss in my piece for DataInformed, virtualization is the starting point for creating this access. Then a semantic layer provides meaning—what we call a semantic metadata catalog.

A data democracy enables access and control. And that's good recipe for innovation.