Six Difficult Questions to Ask About Data Governance

As data continues to pile up, taking full stock and assessing value to it are becoming paramount.

SAN FRANCISCO—With industry analysts warning that business data continues to fill up storage at an average of 60 percent per year, enterprises of all sizes eventually need to make a strategic decision: Will they take better control over the organization of their data as it comes into the system, or will the data simply keep piling up in servers and storage without a clue about access and security?

Steven Adler, IBMs program director of data governance solutions, suggests that being proactive and assessing the notion of "data governance" as soon as possible is a key move for a business of any size.

"Were really talking about organizing human beings to affect certain outcomes. And thats all about behavior," Adler said. "Were coming to this issue of governance because stovepiped management approaches to data and all of its different dimensions arent working."

Adler spoke in keynote addresses last week at both the Data Governance Conference at the Mikayo Hotel here and at the Burton Group Catalyst Conference at the Hilton.

Loosely defined, data governance is managing data as an enterprise asset and controlling operational risk. It means safeguarding corporate information, keeping auditors and regulators satisfied, and using improved data quality to retain customers and drive new business opportunities.

CA, one of the top vendors in the data governance sector, believes that data governance is the bridge between the business people and processes and the IT system of a company. No matter how it is defined, data governance involves work to improve the quality of information stored on databases and typically includes the creation of a team of IT professionals whose sole job is to boost the reliability of the data and improve access to the content.

/zimages/3/28571.gifRead here about Symantecs data risk management tools built into its Information Foundation product.

Among the technologies utilized to support such efforts are software tools used for tracking the manner in which employees are looking at files, and how they behave while logged into databases. CA, IBM, Hewlett-Packard, Oracle, Hitachi Data Systems, Symantec and Scentric are among the market leaders in the sector.

"Were beginning to realize that the category of data is much broader than we thought," Adler said. "This is not just about customer data or employee data, structured and unstructured; its also about video and audio, print, protocols, messages and e-mail—and even software code."

Data is anything that runs on a computer that is created by a person or another computer, Adler said. "All of those different classifications—for different companies and individuals—have value, and also liability," he said. "Management should be concerned with all of it. Thats why we need a governance approach."

Organizations, like governments, have interest groups, Adler said.

"Factions are a way of life for any large organization. You have governance to recognize them and to deal with them," he said. "All these different constituents—whether its security, data quality, data management, data architecture or risk management—all these constituents are key players."

Policy is the articulation of a decision a governing body makes, Adler said. "That could be in the form of a standard, architecture or written policy or rule—or just a memo: Were going to use this vendor as opposed to that one. Thats a policy decision," Adler said.

Sometimes those decisions are made by individuals, but "we hope more often theyre made by groups," Adler said. "Groups seeking to hear different points of view and reach a consensus. The act of sharing common interests and deciding on a common goal—that is the nature of governance."

Adler said every company should ask itself to identify the following six data governance issues:

  • Do you have a government? "Whats its political structure? Who are the players? Are we going to go deep on a few categories, or are we going to go wide on all of them? Companies need to know whos in and whos out [regarding decision-making]," Adler said.

Next Page: Putting a value on data.

Chris Preimesberger

Chris J. Preimesberger

Chris J. Preimesberger is Editor of Features & Analysis at eWEEK, responsible in large part for the publication's coverage areas. In his 12 years and more than 3,900 stories at eWEEK, he...