Data centre analysis and business intelligence guide
A comprehensive collection of articles, videos and more, hand-picked by our editors
Big data is red hot for CEOs, and that is news both good and bad for data management professionals. Jill Dyché, vice president of thought leadership at DataFlux, told delegates at a recent data governance conference in London to be on their mettle, given the heightened visibility data is getting from top business leaders.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
Dyché reported that her Harvard Business Review blog post on how to avoid big data "gotchas" was one of the most popular ever on the website. In that post, she lists five questions executives need to get answers to if they are to commit to big data. There is both burning interest and profound scepticism among C-level leadership, according to Dyché. She said the main question they are asking is, "Does the potential for accelerating existing business processes warrant the enormous cost associated with technology adoption, project ramp up, and staff hiring and training that accompany Big Data efforts?"
For more on potential business value of big data for mainstream organisations
Find out how big data is wading into the mainstream
Gaining business value from big data: a roundtable webcast
Former Yahoo chief Usama Fayyad on how analytics is the future of big data
CEOs are interested in data governance, regulatory compliance, infosecurity, IT modernization, strategic enablement and big data, Dyché said in her summit keynote speech. But the enduring suspicion with big data is that it lends itself merely to academic exercises. And yet, "Companies are more ready than executives know to exploit value of big data."
But, she said: "The ability to use information strategically is a function of the degree to which the data is integrated. The dirty secret of enterprise data warehousing is that the data is often co-located but not integrated.
"Having a big data warehouse does not mean doing big data, which is the evolution of data formats and emerging technologies that are quicker and cheaper. Social media interactions are expensive to do with a data warehouse.
"New technologies can also deal with streaming data from sensors in health care and oil and gas, fast. Show me someone who says they have a decade of experience with big data and I'll show you someone who is misinterpreting their data warehouse!" she added.
While top-level interest in data matters is both welcome and unsettling, so is another trend Dyché identified: lines of business having IT budgets, attenuating centralised IT control. "It means that greater governance than ever is needed of the data."
An important element of that is better classification, using more of a security model. And, in an interview following her summit keynote, Dyché imprecated an enduring disconnect between data governance and infosecurity in large corporate organisations. "These are different tribes. The first thing CEOs think of when you say 'infosec' is stolen laptops. We should think more about policy, who can access what. All data is not the same, and it's not a matter of a whole democracy."
At a DataFlux healthcare client, the head of security has a seat on the data governance council, she said. There needs to be a concept of data regimes, she said, so that some data does not leave its department.
Dyché's advice is to be constantly wary of the "academic exercise" and "start small, think big. Start with small controlled project," proving the business value from there. And, "Don't just grab an executive sponsor. Sometimes executive sponsorship for data governance and master data management is just a tick box."
Even with big data, mainstream organisations "still need to identify a business problem. What's the need, pain or problem? It can't be a research project, as with Google, Facebook, or Yahoo. Even though Hadoop, say, is open source, it is still human intensive.
"And you still need to govern it. 'Why are we doing this? What people are we redeploying in order to do it?' There is a lot of resource questions, a lot of 'why' questions, with big data."