Putting the data governance plan in self-service analytics

As self-service analytics software becomes more common, users are calling on vendors to beef up governance capabilities.

As more and more businesses implement self-service analytics software, a glaring need is becoming apparent: data...

governance plans.

While self-service tools can provide businesses users with a simple way to access and analyze data, they also can give them the ability to feed bad data back into enterprise data stores, silo off important insights or publish information that should be kept private. The need for improved data governance in self-service settings has touched off a race among software vendors to see who can get to the governance finish line first.

In the last two months, both MicroStrategy Inc. and Birst Inc. introduced self-service business intelligence products that stress data governance. MicroStrategy 10 is built around the company's pre-existing enterprise data infrastructure that brings multiple data sources into a common data warehouse accessible by the entire organization. Meanwhile, Birst 5X comes with a data integration layer that can reconcile disparate data types and pull them into an enterprise-wide data warehouse. Visualization tools on top of this essentially let people query data no matter where it is stored.

At the same time, traditional self-service players are touting their governance capabilities more and more. Tableau Software has built in capabilities that allow administrators to assign users to roles that control what they are able to do with data. The QlikView product from Qlik Technologies Inc. comes with a governance dashboard that allows administrators to see which users are using specific data sources and what they're doing with the data.

Increasingly, vendors are seeing their governance capabilities as a differentiator in a crowded market. "That is where the market is going," said Paul Zolfaghari, president of MicroStrategy.

Self-service governance is the expectation

It's not hard to see why vendors are doubling down on their governance functionality. These kinds of features are quickly becoming a minimum requirement, rather than simply "nice to have," as users have already seen how poor governance can lead to siloed and untrustworthy data as well as unauthorized access of protected data.

"The poison of the more open-ended tools would be the inability to be secure," said Jay Egan, senior product manager at The Advisory Board.

Egan and his team use MicroStrategy 10 to deliver data dashboards tracking various quality metrics to healthcare clients. Working in a highly regulated industry in which privacy laws are strong means the abilities to control access to data and ensure its quality are paramount.

Mike Bozek, vice president of business line management at Elekta AB, said he decided to go with the Birst 5X platform primarily because of its data integration capabilities. Elekta is also in the healthcare space, providing electronic health records and performance management systems to hospitals and cancer centers. He said that healthcare as an industry is generally behind the times when it comes to technology adoption, but is digitizing health records and data from patient monitoring systems at a rapid pace. As it does so, it's important to avoid putting in systems that segregate data within departments.

"Most of our clients today use their own Excel spreadsheets or Crystal Reports, but moving to a real-time dashboard is huge," he said.

Evaluating vendors' claims to good governance

Of course, just because vendors are touting their governance plan now doesn't mean they all offer equally strong functionality. Boris Evelson, vice president at Forrester Research, said self-service analytics buyers should look for three distinct capabilities when they're assessing governance:

  • They need to see that a tool allows them to assign specific privileges to users based on predetermined roles.
  • The tool should allow for data to be clearly labeled according to whether it is enterprise data that has been validated or if it is user-generated, and hence, potentially compromised.
  • Self-service tools need to address data preparation. Allowing users to connect to multiple data stores and work with disparate data in one platform will inevitably change the data being used. This process must be controlled.

"Things can get out of control, but there are best practices," Evelson said. "The middle of the road is where you can empower business users to be self-sufficient, but monitor what's going on."

Ed Burns is site editor of SearchBusinessAnalytics. Email him at [email protected] and follow him on Twitter: @EdBurnsTT.

Next Steps

Governance is a key to success with self-services tools

Keep control in self-service BI implementation

Don't overlook governance when putting together self-service projects

Dig Deeper on Self-service and collaborative business intelligence