'Big Data' Prep: 5 Things IT Should Do Now

Got your "big data" plan in place? If not, you may want to start thinking about implementing one.

Big data is being hailed -- or hyped, depending on your point of view -- as a key strategic business asset of the future. That means it's only a matter of time before the suits in the corner office want to know IT's thoughts on the matter.

What to tell them? To be sure, handling large amounts of data isn't virgin territory for most IT departments, but beyond the hype, analysts say, big data really is different from the data warehousing, data mining and business intelligence (BI) analysis that have come before it.

Data is being generated with greater velocity and variability than ever before, and, unlike data in the past, most of it is unstructured and raw (sometimes called "gray data").

Blogs, social media networks, machine sensors and location-based data are generating a whole new universe of unstructured data that -- when quickly captured, managed and analyzed -- can help companies uncover facts and patterns they weren't able to recognize in the past.

"We've collected data for a long time, but it was very limited -- we produced a lot of it, but no one was doing much with it," says Paul Gustafson, director of Computer Sciences Corp.'s Leading Edge Forum, Technology Programs. "The data was archived and it was modeled around business processes, not modeled as a broader set of core knowledge for the enterprise. The mantra is this shift from collecting to connecting."

For example, the U.S. healthcare industry could drive efficiencies and increase productivity by effectively harnessing data related to quality of care, success rates and patient history, according to a May report on big data issued by McKinsey Global Institute, which estimated that the industry could generate more than $300 billion in value every year with such bigdata initiatives. The report likewise suggests that big data has the potential to increase an average retailer's operating margin by more than 60%.

IT is standing at the forefront of this data revolution, industry observers say.

"This is an opportunity to walk into the CEO's office and say, 'I can change this business and provide knowledge at your fingertips in a matter of seconds for a price point I couldn't touch five years ago,'" says Eric Williams, CIO at Catalina Marketing.

Williams should know -- Catalina maintains a 2.5-petabyte customer-loyalty database, including data on more than 190 million U.S. grocery shoppers collected by the largest retail chains. This information is, in turn, used to generate coupons at checkout based on purchase history.

To steer organizations into the era of real-time predictive intelligence, Williams and other industry watchers say, tech managers must evolve their enterprise information management architecture and culture to support advanced analytics on data stores that measure in terabytes and petabytes (potentially scaling to exabytes and zettabytes).

"IT is always saying they want to find ways to get closer to the business -- [big data] is a phenomenal opportunity to do exactly that," Williams says.

Overcoming big-data hurdles

Because it's early on, big-data technologies are still evolving and haven't yet reached the level of product maturity to which IT managers have grown accustomed with enterprise software.

Many emerging big-data products are rooted in open-source technologies, and while commercial distributions are available, many still lack the well-developed third-party consulting and support ecosystem that accompanies traditional enterprise applications like ERP, points out Marcus Collins, research director at Gartner.

What's more, there is a significant gap in big-data skills in most IT departments, which have, up until now, focused on building and maintaining more traditional, structured data warehouses.

And there are major shifts to be made, both in terms of culture and in traditional information management practices, before big data can successfully take hold within an IT organization and throughout the company, notes Mark Beyer, Gartner's research vice president of information management.

Rather than waiting for the pieces to fall into place, savvy IT leaders should start prepping themselves and their organizations to get ahead of the transformation, says Beyer and other analysts.

Here are the top five things tech managers should be doing today to lay out a proper foundation for the big-data era of tomorrow.

Take stock of your data

Nearly every organization potentially has access to a steady stream of unstructured data -- whether it's pouring in from social media networks or from sensors monitoring a factory floor. But just because an organization is producing this fire hose of information, that doesn't mean there's a business imperative to save and act on every byte.

"With this initial surge around big data, people are feeling an artificial need to understand all the data out there coming from weblogs or sensors," notes Neil Raden, vice president and principal analyst at Constellation Research.

Part of that anxiety may be coming from vendors and consultants eager to promote the next big thing in enterprise computing. "There's a certain push to this coming from people who are commercializing the technology," Raden observes.

Smart IT managers will resist the urge to try to drink from the fire hose, and instead serve as a filter in helping to figure out what data is and isn't relevant to the organization.

A good first step is to take stock of what data is created internally and determine what external data sources, if any, would fill in knowledge gaps and bring added insight to the business, Raden says.

Once the data scoping is underway, IT should proceed with highly targeted projects that can be used to showcase results as opposed to opting for a big-bang, big-data project. "You don't have to spend a few million dollars to start a project and see if it's worth it," Raden says.

Next Page: business needs, infrastructure, technology, and staff