In a recent Web seminar entitled “Big Data: Can You Seize the Opportunity” conducted by Harvard Business Review, the basic framework for enterprise big data programs was challenged by presenters Donald Marchand, professor of Strategy Execution and Information Management at IMD in Switzerland and founder of analytics firm enterpriseIQ, and Joe Peppard, analytics consultant and professor of information systems, Cranfield University School of Management in the U.K.
On the enterprise information management front, Peppard and Marchand found the fulcrum for the move to big data projects comes from a trio of changes: a shift from big systems to big data; a change from re-engineering processes to re-engineering decision-making; and a move away from management or control over data to leveraging people-powered data opportunities. Businesses have long held issues in dealing with data, though the attention on big data over the last 18 months has raised the topic among executives and business-side managers as a potential source of returns.
“Today more and more companies are seeking opportunities with more information to more people. This task is quite challenging, and lots of organizations we’ve worked with aren’t really ready for this shift because it has significant cultural implications, as well as personal implications for individual managers,” Peppard said.
More interest in data is a positive, though following through on that for a truly viable big data program requires a change in mindset from the start to avoid all sides involved being bogged down in data sets and let down by immediate returns, according to the two experts. Unlike the traditional “design to build” approach for data projects, business and IT need to come together on the perspective of “design for use” with big data. Contained under this design for use framework is the idea of enterprise information “exploration,” that evolving process you take to frame questions of data and business results. This could dredge up any number of unknown data quality problems – not unlike those found in the process of a mature CRM implementation – and should beat back the notion of big data as a “magic bullet” enterprise project.
It also means that “supplanting IT knowledge with business knowledge becomes critical to understanding how data can be improved and your business’ behaviors with sharing and exchanging information,” Marchand said.
“It means not focusing on outcomes [at the start] but sensing a problem and opportunity along the way, which may change ways which you use information,” he said.
Ultimately, big data requires a modern, direct relationship between technology and business benefit, even without the black-and-white financial rewards stated at the start, Peppard said. These relationships are mediated by change, process and practice, and often face long-standing enterprise information management obstacles between IT and the C-suite. Peppard said that business looks at data almost like “junk food”—they consume what immediately catches their eye for a single point or interest, with little regard to depth or quality of the data involved. And from the IT side of the house, it’s planning that too heavily relies on known existing data and the solutions.
“IT cannot provide a solution for an organizational problem. It might be part of what is required. But while IT can deliver a certain tool or capability, the business itself has got to develop complementary capabilities if the benefits are going to be delivered,” he said.
Justin Kern is senior editor at Information Management magazine, a sister publication of Health Data Management.