The Value of Democratizing Data

By Samuel Greengard Print this article Print
Democratizing Data

There's a growing need to put data to work at all levels of a company. The ability to democratize data increasingly determines a business' success or failure.

The company has focused on collecting a broad spectrum of data and establishing a central data source. Ultimately, "Everyone is tapping into the same data and seeing the same data," she says.

The results have been impressive, particularly for the marketing department. Finish Line, by personalizing communications, has achieved a 50 percent increase in open rates and has increased conversion rates for email, while also realizing a 30 percent increase in gross return on ad spend for Facebook. Meanwhile, various departments and groups are saving time and increasing their efficiency by tapping into the data pool.

"This is where the industry is headed," Bleymaier states. "The ability to use data across an organization is critical."

Creating a New Framework

Building a framework that supports data democratization is not a simple task. E&Y's Schlesinger says that at the most basic level, organizations must build an IT framework that supports a single data repository.

This increasingly means using Hadoop or similar storage environments. It means using clouds strategically, putting APIs to work to connect systems quickly and efficiently, and adopting analytics tools that support enterprisewide use, including on mobile devices. It also means focusing on data lifecycle management, data ownership and governance.

"Even with a tool such as Hadoop, there comes a time when your cluster gets too big and you're not realizing value," Schlesinger points out. "So, you really have to understand data in a broader and deeper context."

IT executives must reevaluate how they approach data management, adds Mark Beyer, Gartner research vice president and distinguished analyst. He says that as organizations transition to more advanced platforms and environments, the focus must be on logical and virtual data warehouses and how to best use these environments. It's also important to look at how to extend virtual access out to a Hadoop cluster, which can run processes through a semantic tier, direct database access or Hadoop software.

Unfortunately, "This new way of architecting a data warehouse is still relatively unknown," Beyer says. "Only about six percent of the entire data warehouse market is aware of this approach and using it."

IT leaders must also consider new ways to architect systems, he points out. One concept Beyer supports involves information message portals. These systems essentially use domain name servers to handle multiple levels of metadata.

This approach, still largely in the conceptual stage, would introduce clearly defined levels of access within an enterprise in order to better match the needs of business units, departments, and other functions or locations, while providing notifications and alerts for changes, as well as new data.

The bottom line? "Today, too much data resides in silos and is unavailable," Beyer states.

Make no mistake: A new era of data management is emerging. As the Internet of things and connected devices and systems take hold, data volumes will continue to expand, and the value of data will increase.

Already, many view data as a form of capital and say that organizations will use it as such in the years ahead. Along the way, the need to democratize data will grow.

"We are entering a new era of data management," E&Y's Schlesinger says. "Business leaders must think about how to manage data throughout the entire lifecycle and make data available to the entire organization as it is needed."

This article was originally published on 2015-07-30
Samuel Greengard writes about business and technology for Baseline, CIO Insight and other publications. His most recent book is The Internet of Things (MIT Press, 2015).
eWeek eWeek

Have the latest technology news and resources emailed to you everyday.