Business groups in a growing number of companies appear to be plowing ahead on data analytics projects with little input or help from their own IT organizations.
Rather than leveraging in-house IT skills and technology, many business groups are using their own data and department-level analysts to cobble together analytics strategies, according to a survey by IDC.
Business managers and IT managers appear to have different assessments of the value enterprise IT organizations bring to big data and data analytics projects. While IT groups see themselves as enablers, business leaders tend to view IT as a stumbling block.
For the study, IDC surveyed 578 line-of-business managers, IT managers, data analysts and business executives.
Close to 38% said the majority of their analytics staff resides outside of IT in a centralized analytics group. Slightly more than 20% said the analytics group is primarily responsible for determining their business unit or department's analytics strategy.
When IT managers were asked whom they thought was primarily responsible for driving analytics efforts, close to four in 10 identified IT as the go-to organization. In contrast, barely 25% of business unit leaders saw IT as best suited to lead analytics strategies.
IT had a similarly rosier view of their organization's available resources for handling analytics projects compared to business managers.
Some 65% of the IT managers surveyed claimed they're either "satisfied" or "very satisfied" with the staff, technology and processes in place for handling analytics projects; a more modest 57% of business managers felt the same way.
A similar dichotomy prevailed around collaborative efforts between IT and business units, with more IT leaders saying they're happy with existing relationships than business managers.
The results bust some of the popular misconceptions that exist around IT's role in enterprise analytics initiatives, said Pamela Prentice, chief research officer at SAS Institute, which commissioned the IDC survey.
The notion that IT controls all data is simply no longer true, Prentice said. "Lines of business are taking a lot of the analytics work for business reasons," she said. Many see IT as being overburdened with operational projects and not responsive enough to business.
"[Business groups] are more enabled. They are getting more authority to go outside of IT and enable their own analytics platforms. They are running things outside of IT," she said.
While many IT groups continue to labor under the impression that they are primarily responsible for their organization's data analytics strategy, business heads see a different role for their technology organizations. Business groups really are looking to IT to be the data monitors responsible for making reliable data available for analytics.
Meanwhile, the actual funding for analytics projects is generally decided by IT and the line of business working together or by the business group on its own. "Few analytics funding decisions are made by IT alone," Prentice said.
The situation presents an opportunity for IT groups to make themselves more relevant, she noted. The trend among business groups to embark on their own analytics projects is leading to the creation of shadow IT organizations, multiple data streams and siloed department and business-level data marts.
IT has an opportunity to make itself more useful in the data analytics and big data arena by taking control of data and making it available in a more manageable form, she said.
Curt Monash, database and analytics expert and principal at Monash Research, said that as companies begin to roll out more enterprise-level analytics efforts IT's role will actually increase, not diminish.
Emerging approaches like predictive data modeling involves dealing not just with large volumes of data, but also with data that is acquired at great speed and stored in different ways. Increasingly, the modeling is done directly against the database itself.
Contrary to what the study suggests, IT has more to contribute to analytics than ever before, Monash said. "IT has a lot of work to do to harness new volumes and velocities and varieties of data."
While smaller, tactical analytics projects may well be handled at the business unit and department level, large strategic analytics projects will lean heavily on IT skills and expertise, he said.
Jaikumar Vijayan covers data security and privacy issues, financial services security and e-voting for Computerworld. Follow Jaikumar on Twitter at @jaivijayan or subscribe to Jaikumar's RSS feed . His e-mail address is email@example.com.