This site uses cookies to provide you with a more responsive and personalized service. By using this site you agree to our use of cookies. Please read our cookie notice for more information on the cookies we use and how to delete or block them.

Bookmark Email Print this page

Finding the Face of Your Data

Hidden patterns in big data can trigger breakthrough insights – if you have the data scientists to guide discovery

Back
Next
PC_01
MB_02
SO_03
DE_04
IP_05
FD_06
GM_07
EE_08
HP_09
IT_10
IT_10


Download chapter  Watch video  Read more about Big Data

The majority of big-data market growth can be attributed to business’s increasing interest in related analytics capabilities to drive improved decisions through improved insight. Not surprisingly, some business executives feel big data is over-hyped. Others claim its value is under-realized.

Big data alone, however, creates no new value if it doesn’t lead to insights – about questions you haven’t answered before, or perhaps more importantly, about questions you didn’t know you could ask. For that, you’ll need new ways to capture and explore data, to enrich data and to interact with data. You should also identify new patterns in areas you’ve never explored, anticipate potential value and insights, and zero in on specific crunchy questions whose answers can lift performance and competitiveness. Man and machine, working together, to find the face of your data.


My Take
Hear Tom Soderstrom, IT Chief Technology Officer at NASA’s Jet Propulsion Laboratory, describe first-hand his perspective on Finding the Face of Your Data.

 

Watch video

Mark White, principal and chief technology officer, Deloitte Consulting LLP, describes how big data and sophisticated visualization technologies are changing the way some pharmaceutical companies approach clinical trials.

 

Read more about Finding the Face of Your Data

Where do you start?

Information has been the white whale of the last decade of IT investment, but few organizations have mastered core data management across the enterprise – much less higher-order disciplines like business intelligence, visualization, or advanced analytics. Does that mean you shouldn’t press forward? Hardly.

Coupling data scientists and machine learning with visualization and other data interaction tools can allow timely results with less infrastructure build-out – and gives a higher tolerance for error. Imperfections of data sources and the limitations of machine learning are hedged by intuition and cognitive insights of human counterparts. Technology advances allow for data crunching that would have crippled sophisticated servers just five years ago. Some specific areas to start the journey:

  • Stack the deck. Don’t play without talent. Start with business domain experience – people to help inventory and categorize the information assets at your disposal, understand line-of-business priorities and work with technology specialists to determine short- and long-term needs.
  • Look to the source. Finding the Face of Your Data begins and ends with data sources. The problem is that those sources are often an unknown quantity. You should understand not only your own data, but also the third- party sources available to supplement and extend your own purview. Open source, governmental and subscription-based feeds may provide crucial missing links to discover new patterns and insights.
  • Start the clock. Big data projects can spiral in complexity and scope. Identifying sources, integrating the data, massaging the data, running models and creating visualizations and reports can exhaust an enthusiastic business sponsor’s appetite. Start small – delivering something quickly before the business users lose interest.
  • Manage expectations. Trying to launch this effort with enterprise-wide plans would likely be daunting. Choose specific domains in which to begin (e.g., customer, product, pricing, risk). Focus on places where your available people are most knowledgeable. Instead of dealing with hypothetical super-sets, focus on tangible, bounded focus areas. Run some experiments to test the data and deliver results. But once the scope has been anchored, don’t limit yourself to just one domain. Investigate seemingly unrelated data sets. You never know where an “aha” relationship may be lurking.
  • Good hygiene. Big data is still data, and you’ll still need data disciplines. Core data management, master data management, integration and stewardship are important – even if only for the small domain slices upon which early initiatives are sighted. This is the structural foundation for future efforts. Enough of a strategic roadmap should be laid out to monitor that short-term enablement is in line with long-term vision.
  • Stay on target. Before you can determine the right questions to ask, you’ll need to illuminate the possible questions that could be answered. Then define relative priorities and associated metrics to assess what changes could be enacted based on insights. Early focus should be on areas that involve no more than first- or second-degree inferential analysis.

Bottom line

Information as an asset is moving to the forefront of the business agenda. From basic analytics to sophisticated visualization. From taking advantage of big data to pattern discovery and sense-making. Despite many advances, we remain a long way from completely automated discovery and contextual analysis. Machines are very good at finding patterns, but we should still have human analysts – augmented by visualization tools – to determine real meaning and value.

In Finding the Face of Your Data, we are building on the foundational elements of previous cycles and trends, while recognizing the importance of balancing people, data, and computing power to discover questions that previously couldn’t be answered – or even asked. The goal remains the same: data-driven decision making, with higher speed-to-value.

 

Related links

Share this page

Email this Send to LinkedIn Send to Facebook Tweet this More sharing options

Stay connected